Frequency converter - Doria [PDF]

LAPPEENRANNAN TEKNILLINEN YLIOPISTO Faculty of Technology Energy Technology MASTER’S THESIS

CALCULATION ANALYSIS OF ENERGY SAVING TOOLS FOR FAN AND PUMP APPLICATIONS

Examiners:

Professori Esa Marttila TkL Simo Hammo

Supervisor:

M.Sc. Jukka Tolvanen

Lappeenranta x.x.2008 Teemu Taskinen Huhtiniemenkatu 12 as 6 53810 Lappeenranta Gsm. +358 40 568 1696

ABSTRACT Lappeenranta University of Technology Faculty of Technology Energy technology Teemu Taskinen Calculation analysis of energy saving tools for fan and pump applications Master’s thesis 2008 189 pages, 92 figures, 31 tables and 7 appendices Examiners:

Professor Esa Marttila Lic.Sc. (Tech.) Simo Hammo

Keywords:

FanSave, PumpSave, fan, pump, frequency converter

This thesis analyses the calculation of FanSave and PumpSave energy saving tools calculation. With these programs energy consumption of variable speed drive control for fans and pumps can be compared to other control methods. With FanSave centrifugal and axial fans can be examined and PumpSave deals with centrifugal pumps. By means of these programs also suitable frequency converter can be chosen from the ABB collection. Programs need as initial values information about the appliances like amount of flow and efficiencies. Operation time is important factor when calculating the energy consumption and information about it are the length and profile. Basic theory related to fans and pumps is introduced without more precise instructions for dimensioning. FanSave and PumpSave contain various methods for flow control. These control methods are introduced in the thesis based on their operational principles and suitability. Also squirrel cage motor and frequency converter are introduced because of their close involvement to fans and pumps. Second part of the thesis contains comparison between results of FanSave’s and PumpSave’s calculation and performance curve based calculation. Also laboratory tests were made with centrifugal and axial fan and also with centrifugal pump. With the results from this thesis the calculation of these programs can be adjusted to be more accurate and also some new features can be added.

TIIVISTELMÄ Lappeenrannan teknillinen yliopisto Teknillinen tiedekunta Energiatekniikka Teemu Taskinen Pumppu- ja puhallinlaitteistojen energiansäästötyökalujen laskennan analysointi Diplomityö 2008 189 sivua, 92 kuvaa, 31 taulukkoa ja 7 liitettä Tarkastajat:

Professori Esa Marttila TkL Simo Hammo

Hakusanat: Keywords:

FanSave, PumpSave, puhallin, pumppu, taajuusmuuttaja FanSave, PumpSave, fan, pump, frequency converter

Diplomityön tarkoituksen on tarkastella energiansäästötyökalujen, FanSave ja PumpSave, laskentaa. Ohjelmilla voidaan tarkastella puhaltimien ja pumppujen taajuusmuuttajakäytön energian kulutusta ja vertailla sitä muihin säätötapoihin. FanSave –ohjelmalla voidaan käsitellä keskipakoisja aksiaalipuhaltimia ja PumpSave –ohjelmalla keskipakoispumppuja. Ohjelmien avulla voidaan myös valita tarkasteltavalle puhaltimelle ja pumpulle sopiva taajuusmuuttaja ABB:n mallistosta. Ohjelmiin täytyy syöttää alkuarvoina tietoja laitteistosta, kuten virtauksen määrä ja hyötysuhteita. Laitteiston käyttöaika on tärkeä tekijä energian kulutusta laskettaessa ja sille voidaan määrittää kokonaiskesto ja profiili. Diplomityössä esitellään puhaltimien ja pumppujen toimintaan liittyvä perusteoria ilman tarkempia selvityksiä mitoitusperiaatteista. FanSave ja PumpSave –ohjelmat sisältävät monia eri säätötapoja virtaukselle. Nämä säätötavat esitellään työssä niiden toimintaperiaatteiden ja soveltuvuuden osalta. Myös puhaltimien ja pumppujen käyttöön olennaisesti liittyvät oikosulkumoottori ja taajuusmuuttaja esitellään työssä. Työssä vertaillaan energiansäästöohjelmien laskentaa puhaltimien ja pumppujen ominaiskäyrästöjen perusteella laskettuihin tuloksiin. Vertailupohjaa ohjelmien laskentaan saatiin myös suorittamalla laboratoriotestejä keskipakois- ja aksiaalipuhaltimilla sekä keskipakoispumpulla. Työssä saatujen tuloksien perusteella ohjelmien laskentaa voidaan säätää tarkemmaksi sekä lisätä siihen ominaisuuksia.

TABLE OF CONTENTS 1. INTRODUCTION ..............................................................................................................5 1.1 Background & objectives..............................................................................................5 1.2 Execution of the work...................................................................................................5 2. FANS ..................................................................................................................................6 2.1 Structure and operation of a centrifugal fan .................................................................6 2.1.1 Forward curved blades...........................................................................................7 2.1.2 Backward curved blades ........................................................................................8 2.1.3 Radial blades..........................................................................................................8 2.2 Structure and operation of an axial fan .........................................................................9 2.2.2 Axial fan with fixed blades..................................................................................10 2.2.3 Adjustable axial fans............................................................................................10 2.3 Fan theory ...................................................................................................................13 2.3.1 Specific fan power ...............................................................................................18 2.3.2 SFS ISO 5167 standard........................................................................................18 2.3.1 Characteristic curves for a fan .............................................................................21 2.4 Fan flow control methods ...........................................................................................21 2.4.1 No flow control....................................................................................................22 2.4.2 Rotational speed control ......................................................................................23 2.4.3 Outlet damper ......................................................................................................24 2.4.4 Slip coupling ........................................................................................................25 2.4.5 Voltage control ....................................................................................................25 2.4.6 Two speed motor .................................................................................................26 2.4.7 Cycling on/off ......................................................................................................26 2.4.8 Inlet box damper ..................................................................................................26 2.4.9 Inlet vanes for centrifugal fans ............................................................................27 2.4.10 Pitch angle adjustment in axial fans ..................................................................28 3. CENTRIFUGAL PUMPS.................................................................................................28 3.1 Structure and operation of centrifugal pump ..............................................................29 3.2 Pump hydraulics .........................................................................................................30 3.2.1 Pump power demand and efficiency....................................................................30 3.2.2 System head and net positive suction head..........................................................31

3.2.3 Performance curves and affinity laws..................................................................34 3.3. Flow control methods in pumping .............................................................................36 3.3.1 Variable speed control .........................................................................................37 3.3.2 Throttling control.................................................................................................38 3.3.3 On/off –control ....................................................................................................39 3.3.4 Other control methods .........................................................................................39 4 SQUIRREL GAGE MOTOR ............................................................................................40 4.1 Structure of squirrel gage motor .................................................................................41 4.2 Motor losses................................................................................................................41 5. FREQUENCY CONVERTER .........................................................................................42 5.1 Drive losses.................................................................................................................44 6. FANSAVE AND PUMPSAVE ENERGYSAVING TOOLS..........................................44 6.1 FanSave.......................................................................................................................45 6.1.1 Initial values.........................................................................................................45 6.1.2 Calculations .........................................................................................................46 6.2 PumpSave ...................................................................................................................50 6.2.1 Initial values.........................................................................................................50 6.2.2 Calculations .........................................................................................................51 7. COMPARISON BETWEEN FANSAVE’S CALCULATION AND PERFORMANCE CURVE BASED CALCULATION .....................................................................................55 7.1 Centrifugal fans...........................................................................................................56 7.1.1 Radial blades........................................................................................................57 7.1.2 Fans with forward curved blades. ........................................................................62 7.1.3 Fans with backward curved blades ......................................................................67 7.2 Axial fans ....................................................................................................................76 8. LABORATORY TEST RUNS WITH CENTRIFUGAL FAN........................................89 8.1 Laboratory equipment.................................................................................................89 8.2 Measurements and comparison of results to FanSave’s calculation...........................91 8.2.1 Outlet damper ......................................................................................................91 8.2.2 Variable speed control .........................................................................................92 8.2.3 Two speed motor .................................................................................................94 8.2.4 Inlet box damper ..................................................................................................94 8.3 Energy consumption examination in operation time ..................................................96 9 LABORATORY TEST RUNS WITH AN AXIAL FAN .................................................96 9.1 Laboratory equipment.................................................................................................97

9.2 Variable speed drive ...................................................................................................97 9.3 Outlet damper .............................................................................................................98 9.4 Inlet damper ................................................................................................................99 10 PERFORMANCE CURVE BASED CALCULATION COMPARISON TO PUMPSAVE’S CALCULATION ......................................................................................101 10.1 VSD ........................................................................................................................101 10.2 Throttling ................................................................................................................115 11 LABORATORY TEST RUNS WITH A PUMP...........................................................129 11.1 VSD ........................................................................................................................130 11.2 Throttling ................................................................................................................133 12 SUMMARY...................................................................................................................137 12.1 Observation of the profitability of investment........................................................138 APPENDICES

3

SYMBOLS AND ABBREVIATIONS

Greek letters

β

Diameter ratio

ε

Expansibility factor

λ

Friction-coefficient

ρ

Density

ζ

Loss-coefficient

η

Efficiency

κ

Isentropic exponent

μ

Dynaaminen viskositeetti

υ

Kinetic viscosity

ω

angular speed

[kg/m³]

[Ns/m]

[rad/s]

Roman letters A

Area

[m²]

C

Discharge coefficient

c

Heat capacity

[J/kgK]

D

Pipe diameter

[m], [mm]

E

Energy

[kWh]

d

Orifice plate diameter

[m], [mm]

g

Gravitational factor

[m/s²]

h

Enthalpy

[J/kg]

l L

quotient of a distance

n

Rotational speed

[rpm]

NPSH

Net Positive suction head

[m]

P

Power

[W], [kW]

p

Pressure

[Pa]

q

Flow rate

[m³/s], [kg/s]

R

Universal constant for gas

[J/kgK]

r

Radius

[m], [mm]

4

T

Temperature

[K]

u

Tangential speed

[m/s]

v

Flow velocity

[m/s]

Subscript 0

initial

1

inlet

2

outlet

e

electricity

i

flow amount

m

mass

r

friction

s

isentropic

v

volume

vsd

variable speed drive

Acronyms SFP

Specific fan power

VSD

Variable speed drive

5

1. INTRODUCTION Industry uses lots of different types of pumps and fans in needs of different processes. These devices run long times at fast speeds. All this needs lots of energy and use of pumps and fans cover a significant part of industrys electricity bill. Therefore it is essential that the use of pumps and fans is done as efficiently as possible. Flowing in pumps and fans can be controlled in many different solutions but variable speed drive has begun to become more and more common. This is caused by advancements in technology and cost reduction in investment. With variable speed drive pumps and fans can be used precisely matching the needs of the process. Therefore the whole process can be controlled more accurately. Also the wear and noise on machinery is reduced. By using variable speed drive energy can be saved and therefore also the emissions are reduced. In favour of variable speed drive are also constantly increasing price of electricity and tightening emission limits.

1.1 Background & objectives The purpose of this thesis is to analyse the calculations of two Excel based energy saving tools called FanSave and PumpSave by ABB Oy. With these tools the different flow control methods for pumps and fans can be compared. These programs calculate the costs of different flow control methods compared to variable speed drive. For fans and pumps basic theory is introduced but any specific dimensioning of devices is not processed because either FanSave or PumpSave do not include or demand so specific information. Also parallel connection or series connection of fans and pumps is not dealt with in this thesis because FanSave and PumpSave examine only one application at a time.

1.2 Execution of the work Thesis starts with introduction to structure and operation of fans and pumps. Also the basic theory of pumping and blowing is introduced. With FanSave axial and centrifugal fans can

6

be examined so these are the fan types studied in this thesis. PumpSave deals only with centrifugal pumps so other pump types were left aside. Flow during pumping and blowing can be adjusted in many different ways and the most common ones of these are examined and compared to FanSave’s and PumpSave’s calculation in different ways. All the other flow control methods and their operation principles are also introduced. Next part of the thesis consits of introduction of energy saving tools FanSave and PumpSave. Calculations of both of the programs are presented in equations. Results from the calculations of these programs are compared to calculations based on different types of fans and pumps performance curves and to tests performed in laboratory. Laboratory tests were performed in laboratories of Lappeenranta university of technology. All the flow control methods included in the programs could not be tested due to limitations of laboratory equipment and rareness of control methods. Results are presented in tables and as curves for clarification. Thesis ends with conclusions and proposals for improvements and corrections for the energy saving tools.

2. FANS Fan is the most important part of a ventilation system. It is used to transfer air through ducts and spaces. Air movement is caused by a rotating impeller which is installed with blades. Fans are chosen by the purpose of use. Most important factors affecting the choice are the volume flow and pressure difference generated by the fan. For the most powerful fans also the changes in density and temperature of the fluid should be noticed when estimating the power demand of the fan. (Seppänen, 121)

2.1 Structure and operation of a centrifugal fan Flow enters the centrifugal fan in the direction of the axel and leaves the fan at perpendicular to the axel. Three most common types of blades are forward curved blades, backward curved blades and radial blades. (Seppänen, 121)

7

2.1.1 Forward curved blades Characteristic for an impeller of a fan which has forward curved blades is that the impeller is small-sized and has low rotational speed and noise level for certain pressure and volume flow. Change in characteristic curve of a fan causes only a small change in pressure difference between inlet and outlet of a fan. Solid matter particles can easily stuck to a blade which is curved forward and that is why these types of blades are not used for blowing combustion gases. (Puhaltimet teollisuudessa ja LVI-laitoksissa, II 1, Fan Types) Pressure and power demand curve of a centrifugal fan with forward curved blades are illustrated in figure 1. At low flow rates the operation of fan is unstable this is shown in pressure curve which is descending at this operation area. Using fan in this operation area is not suitable but at the operation range where pressure curve is ascending. Instability of the system is shown also in the section angle of the system curve and the pressure curve which is small. If high pressure generation is needed without changing the radial velocity forward curved blades are a good option. Fan with forward curved blades can reach the efficiency of about 60 %. (VLV-kone luentomoniste 1, s 67 ; Systemair)

Figure 1. Pressure curve for a centrifugal fan with forward curved blades. (VLV-kone luentomoniste 1, sivu 68)

8

2.1.2 Backward curved blades Impeller which has backward curved blades it is characteristic to have high efficiency and higher noise level than an impeller with forward curved blades. When fan’s characteristic curve changes the volume flow doesn’t really change. Impeller with backward curved blades is suitable for cases where low running costs are demanded. This type of impeller is also sensitive for gathering particles from gas. For energy efficiency the backward curved blades are better choice than forward curved blades or radial blades. (Puhaltimet teollisuudessa ja LVI-laitoksissa, II 1, Fan Types) Figure 2 illustrates pressure and power demand curves for a fan with backward curved blades. This type of fan is able to function at wide operation range. Fan with backward curved blades can reach the efficiency of 80 %. Power demand is at its highest at the same area the efficiency is at its highest. (VLV-kone luentomoniste 1, 67 ; Systemair)

Figure 2. Pressure curve for a centrifugal fan with backward curved blades. (VLV-kone luentomoniste 1, s 68)

2.1.3 Radial blades

9

Fan with radial blades is the best choice for blowing gas which has lots of solid particles because radial blades are better to resist particles sticking on them than curved blades. Radial blades are straight pointing directly to the ring of the impeller. Fan equipped with radial blades can gain maximum efficiency of 55 %. (Fan Types, Systemair)

2.2 Structure and operation of an axial fan Axial fan is suitable for many applications because it is easy to install and it is small-sized and cheap. Characteristic for axial fan is big volume flow with moderate pressure development. Two phased axial fan has become more general due to its better pressure development abilities. Two phased axial fan consists of two series-connected fans which rotate to opposite directions. This solution generates almost three times as much pressure as a single axial fan. When volume flow of air is big axial fan is a suitable choice for almost any general ventilation applications. Pressure generation in fan can be squared in function of impeller diameter. (Puhaltimet teollisuudessa ja LVI-laitoksissa, I 2-3, 10) Erään aksiaalipuhaltimen, jonka potkurin siipikulma on suuri, painekäyrä on kuvassa 3. Kuvassa nähdään sakkauskohta, jossa tehontarve käyrä saavuttaa huippukohdan. Puhaltimen käynti sakkauskohdan vasemmalla puolella on epävakaata. Puhallinta tulisi käyttää alueella, jossa painekäyrä on nouseva. Kuvassa 3 näkyvä tehontarvekäyrä on riippuvainen impellerin siipiprofiilista eikä siksi päde kaikille aksiaalipuhaltimille. Joillakin puhaltimilla tehontarve voi olla pienellä virtauksella lähes kaksinkertainen sakkauspisteen

tehontarpeeseen

verrattuna.

Tällöin

on

vaarana

käyttömoottorin

ylikuormittuminen. Aksiaalipuhaltimen hyötysuhde on normaalisti maksimissaan n. 75 %, mutta johtosiivillä se voidaan nostaa n. 85 %:iin. (VLV-kone luentomoniste 1, s 67)

10

Figure 3. Pressure curve for an axial fan. (VLV-kone luentomoniste 1, s 68)

2.2.2 Axial fan with fixed blades Axial fan with fixed blades is suitable for duct installation. Lack of flow adjusting abilities limits the fan’s suitability to applications where power consumption and noise level are not important selection criteria. (Puhaltimet teollisuudessa ja LVI-laitoksissa, I 1)

2.2.3 Adjustable axial fans There are types of axial fans which blade angle can be adjusted during operation or during downtime. With downtime adjustment the adjustment needs the fan to be stopped but when the fan operates at its optimum operation area its power consumption is small. It is suitable for any industrial ventilation applications where axial fan’s pressure development abilities are sufficient Some axial fans have the possibility to adjust blade angles during operation. With this type of adjustment the pressure and volume flow values can be rapidly changed to match processes needs. Adjustment can be done with power or compressed air. This type of fan is suitable for applications where maximum volume flow or maximum pressure is needed only momentarily. (Puhaltimet teollisuudessa ja LVI-laitoksissa, I 1-2)

11

Centrifugal and axial fan have their own characteristic operational areas where they are more suitable than other types of fans. Figure 4 illustrates the operational ranges of the most common fan types. Uppermost operation area belongs to centrifugal fan which is suitable for smaller volume flows but then again for higher volume flows and higher pressure differences. Middle operation area belongs to two phased axial fan which can produce average pressure difference in high flow rates. Lowest operation area belongs to one phase axial fan. In practice these operational ranges can not be divided as clearly as in figure 4 but they overlap each other more or less. Also in some situations some other fan is more suitable to be used in some others nominal operation range.

Figure 4. Characteristic operation areas for different types of fans. (VLV-kone luentomoniste 1, s 68)

Typical efficiency curves for different types of fans are illustrated in figure 5. Total efficiency of the fan means the efficiency calculated by the theoretic power demand which is calculated by using the total pressure generated by the fan compared to the power provided to the impeller. Efficiencies calculated for the curves in figure 5 are average values from different types of fans so they may seem a little low. Upmost dash line represents the efficiency of a centrifugal fan with backward curved blades and axial fan. Lower dash line illustrates the efficiency of a centrifugal fan with forward curved blades. Curve does not reach the highest volume flows because this type of fan is not usually

12

manufactured for such operation areas. Solid lines in figure 10 illustrate total efficiency of blowing. This means how many percentages of electric power from the network can be transformed into motion and pressure of fluid. Efficiency of the motor is presented as second to upmost line. (VLV-kone luentomoniste 1, 68-69)

Figrue 5. Efficiencies of different types of fans. (VLV-kone luentomoniste 1, s 69)

Centrifugal fan’s impellers can be divided into three groups according to curvature of blades. Blades can be bent backward or forward or they can be radial. Due to curvature the wheels are described as F-wheel, B-wheel and R-wheel. Figure 6 illustrates these different solutions for fan impellers. First impeller from the left is impeller with backward curved wheels where angle β 2 is below 90°. One in the middle is impeller with radial blades. Blade angle equals 90°. Last one illustrates forward curved blades and the blade angle is over 90°. The best efficiency can be gained with backward curved blades as radial blades give the most power. Forward curved blades are the best choice if maximum pressure generation is necessary. (Puhaltimet teollisuudessa ja LVI-laitoksissa, II 1, VLV-kone luentomoniste 1, s 57)

13

Figure 6. Different shapes of fan impeller blades and velocity triangles. (VLV-kone s 57)

2.3 Fan theory Many factors affects on power demand of a fan. According to equation 1 main factors are volume flow and pressure difference between inlet and outlet of the fan. During blowing density and temperature of gas changes and these factors must be noticed too. Following part of this thesis deals with theory involved with fans. Isentropic power demand for a fan can be calculated as equation 1 shows. In this case the density of the gas is constant. Ps = Δpq v

(1)

where Ps

isentropic power demand [W]

Δp

pressure rise in fan [Pa]

Isentropic power demand can be expressed as shown on equation 2 when density change in gas has been noticed.

Ps = Δp

qm ρ

where qm

mass flow [kg/s]

(2)

14

Average value of density can be calculated with equation

ρ=

ρ 0 + ρ1 2

(3)

where

ρ0

density of the gas before the fan [kg/m³]

ρ1

density of the gas after the fan [kg/m³]

Change in enthalpy can be calculated as equation 4 or equation 5 shows.

Δh =

Δp

(4)

ρη s

where

ηs

isentropic efficiency

Common value for isentropic efficiency for fans vary from 0,6 to 0,85.

Δh =

Δh s

(5)

ηs

where Δh s

isentropic change in enthalpy [kJ/kg]

Common theory of fluid machinery is also valid for fans. According to this isentropic change in enthalpy can also be presented with equation

(

Δh s = η h c u 2 u 2 − c u1 u 1

)

where

ηh

hydraulic efficiency

(6)

15

c u1

tangential component for absolute speed at inlet edge [m/s]

c u2

tangential component for absolute speed at discharge edge [m/s]

u1

tangential speed at inlet edge [m/s]

u2

tangential speed at discharge edge [m/s]

Components c u1 and u 1 are 0 when the air sucked into the fan is for example outside air and air is not blown into the inlet of the fan. Component c u 2 depends on the shape of the impeller blades, the blade angle β 2 and the tangential speed u 2 as shown in figure 9. Hydraulic efficiency contains leakage from discharge outlet to suction inlet, secondary twirls and wall friction. Angular speed for fan can be calculated as equation 7 shows when rotation speed of the fan is known. ω = 2π n

(7)

where ω

angular speed [rad/s]

n

rotation speed of the fan [rpm]

Tangential speed at discharge edge of the impeller can be calculated as equation 8 shows when radius of the impeller is known. u 2 = ωr2

(8)

where r2

radius of the impeller [m]

Gas warms up when it flows through a fan because losses at the flow through the impeller are transformed into heat. How much gas warms up can be calculated by marking the loss power and the heat power equal.

16

Ps

ηm

(1 − η m ) = ρ 0 c p q v Δt → Δt = Ps (1 − η m ) ηm ρ0cpq v

(9)

where ηm

mechanic efficiency of the fan

cp

specific heat capacity for gas [kJ/kgK]

Δt

temperature rise [°C]

Because the power transferred to gas is product of pressure rise and volume flow can equation 9 be written as equation

Δt =

Δp ρ 0 c pη m

(10)

Temperature after the fan can also be calculated as equation 11 shows if the enthalpy change in the fan can be calculated.

T1 = T0 +

Δh cp

(11)

where T0

gas temperature before the fan [K]

T1

gas temperature after the fan [K]

Density of the gas can be resolved as equation 12 shows when temperature and pressure are known.

ρ=

where

p RT

(12)

17

R

universal constant for gas [kJ/kgK]

Actual power transferred to gas can be calculated with equation

P =

Δpq m ρη s

= Δhq m = ΔTc p q m

(13)

where P

power transferred to gas [W]

Fan’s shaft power demand can be calculated with equation

Paks =

P ηm

Paks

shaft power demand [W]

(14)

where

Mechanic efficiency or the fan contains losses in bearings and transmission. Demand for electric power can be resolved from equation

Pe =

P ηm ηe

(15)

where Pe

electric power demand [W]

ηe

efficiency of the motor

Total efficiency of the fan can be calculated with equation η = ηs η m ηe

(16)

18

where

η

total efficiency

2.3.1 Specific fan power

Comparison of fans is made easier with SFP number which means specific fan power. It defines how much power is needed to move a cubic metre of air. In other words it determines the efficiency of air ventilation system regarding to the needed electric power. For a single fan SFP can be expressed with equation

SFP =

Ptotal qv

SFP

specific fan power [kW/(m³/s)]

Ptotal

electric power input to fan application [kW]

(17)

where

Term Ptotal contains the power demand of frequency converter or some other adjusting device as well as the power demand of the driving motor. So it takes into account all the losses in application. For ventilation systems with multiple fans the SFP number is calculated otherwise. (SFP-opas)

2.3.2 SFS ISO 5167 standard

Standard SFS ISO 5167 was applied for laboratory test results from centrifugal and axial fan test runs. Standard was used to calculate volume flow when observing fan’s power demand in different flow control methods. ISO 5167 standard contains instructions to solve different quantities with iteration. SFS ISO 5167 standard can be used for calculating different factors of flow. Mass flow measured with an orifice plate the mass flow can be calculated with equation

19

qm =

C 1− β

4

ε

π 4

d 2 2Δpρ

(18)

Where

β

diameter ratio

Δp

pressure difference over the orifice plate

ε

expansibility factor

C

discharge coefficient

d

diameter of the hole in the orifice plate

Volume flow can be calculated by using density of the fluid as equation 19 shows. q v = ρq m

(19)

Discharge coefficient C is calculated with equation

⎛ 10 6 β C = 0,5961 + 0,0261β − 0,216β + 0,000521⎜⎜ ⎝ Re D 2

+ (0,0188 + 0,0063A )β 4 (1 − 0,11A ) β 4

1− β

8

3, 5

⎛ 10 6 ⎜⎜ ⎝ Re D

(

⎞ ⎟⎟ ⎠

0,3

(

⎞ ⎟⎟ ⎠

0, 7

)

+ 0,043 + 0,080e −10 L1 − 0,123e −7 L1 *

− 0,031 M '2 − 0,8M '2

1,1

)β

(20)

1, 3

Where L1

quotient of the distance of the upstream tapping from the upstream face of the plate and the pipe diameter

⎛ 19000β ⎞ ⎟⎟ A = ⎜⎜ ⎝ Re D ⎠

0 ,8

Term M '2 in equation 20 is calculated with equation

(21)

20

M 2' =

2 L'2 1− β

(22)

Where L'2

quotient of the distance of the downstream tapping from the downstream face of the plate and the pipe diameter

The orifice plate used has corner tappings so equation 23 is valid L1 = L'2 = 0

(23)

Expansibility factor ε is calculated with equation 1 ⎤ ⎡ κ ⎛ ⎞ p 2 ⎢ ε = 1 − 0,351 + 0,256β + 0,93β 1 − ⎜⎜ ⎟⎟ ⎥ ⎢ ⎝ p1 ⎠ ⎥ ⎥⎦ ⎢⎣

(

4

8

)

(24)

When p2 ≥ 0,75 p1

Where

κ

isentropic exponent

p1

pressure before the orifice plate

p2

pressure after the orifice plate

In the iteration the unchangeable factors of equation 18 are marked as A 1 .

A1 =

εd 2 2Δpρ1 μ1 D 1 − β 4

Changing factor in the iteration is marked X 1 .

(25)

21

X 1 = Re D = CA 1

(26)

For discharge coefficient C at the first round of iteration is recommended to use value C ∞ which can be found from table in appendix C. Now Reynolds’s number can be calculated with equation 27. With factor X 1 the first value to volume flow can be solved with equation

qm =

π 4

μDX1

(27)

Second round in iteration is started by calculating value for the discharge coefficient C with the Reynolds’s number. Iteration continues by calculating the value for factor X 1 . Iteration rounds are repeated until the value for volume flow has converged enough.

2.3.1 Characteristic curves for a fan

Fan’s characteristic curve is presented as pressure difference in the function of volume flow. Characteristic curve can also contain curves or areas for fan’s efficiency, power demand, rotation speed and noise level. With the characteristic curve operational values of a fan can be solved at different operation points and in different control methods. Throttling of the flow with dampers and rotational speed control are the most commonly used control methods for volume flow. When flow is throttled changes in fan’s operation can be solved by moving through fan’s curve to a new operation point. With rotation speed control the moving is performed through the system curve. Characteristic curves for different types of fans can be found from appendixes.

2.4 Fan flow control methods Adjusting the gas flow in fans to match the needs of process saves energy. Throttling was most commonly used control method when energy was cheap but nowadays technology

22

has had to come up with new solutions to control air flow. Figure 7 illustrates dependencies of power demand to volume flow in different flow control methods. Fans compared are operating near their best efficiency with axel power ranging from 20 to 30 kW. At the total pressure 800 Pa this means that volume flow ranges from 20 to 30 m³/s. Resistance curve of the duct is assumed to be a second-order parable. In figure 18 curve 1 illustrates a lossless power transferred to gas in rotational speed control, curve 2 illustrates a centrifugal fan with rotational speed control, curve 3 illustrates an axial fan with pitch angle adjustment, curve 4 illustrates a centrifugal fan with inlet vanes, curve 5 illustrates a centrifugal fan with throttling control and curve 6 illustrates an axial fan with throttling control. (Seppänen, 146-147)

Figure 7. Dependency of power demand to volume flow. (Seppänen, 147)

2.4.1 No flow control

This flow control method is used only with small applications and it can be used with many fan types like propeller fans, axial fans and centrifugal fans. The impeller of the fan is installed on the axel of the motor. Therefore there is no transmission loss. The motor is exposed to the gas going through the fan so it must endure all the temperatures of the gas.

23

Rotational speed of the fan can not be changed in this flow control method so it is very important to choose the right type of fan for the application. (Seppänen, 142)

2.4.2 Rotational speed control

Applications with small power consumption speed control are carried out with the use of thyristors. Larger applications use frequency converters. Thyristor based speed control is cheap to implement but the receivable savings are not significant since motor power is small. Thyristor based speed control is feasible for applications such as heat extraction where gas flow and pressure generation through fan are small. Other control methods for rotational speed are squirrel gage motor, which has several windings for different rotational speeds, using a v-belt drive and changing belt pulleys, using a transmission and using a hydraulic clutch. (Puhaltimet teollisuudessa ja LVI-laitoksissa, I 7, Seppänen, 146) Use of frequency control can produce significant saving in energy because the volume flow of the gas can be exactly matched with process’ needs. Other benefits of using variable speed drive are that noise level drops when revs drop and that fan efficiency stays high at the whole control range. Frequency converters have been used for bigger applications for some time now but are becoming more and more general also in smaller applications. Figure 8 illustrates how fan’s operational values change when rotation speed changes. (Puhaltimet teollisuudessa ja LVI-laitoksissa, I 7-8 II 6, Seppänen, 146)

24

Figure 8. Effects on characteristic curves of a fan when rotation speed changes. (Seppänen, 146)

2.4.3 Outlet damper

Outlet damper means flow control method where adjustable dampers are installed after the fan to control the volume flow. This method is simple and cheap to implement but it increases flow resistance in the duct and can cause noise level increase. Outlet damper as a flow control method can be observed as a throttling control. (Puhaltimet teollisuudessa ja LVI-laitoksissa, II 5) Figure 9 illustrates the effect outlet damper flow control method has on the operation point of the fan. At operating point A the dampers are fully open and at point B dampers start to close. Closing the dampers produces always a new system curve when the operating point moves to the intersection point of the pressure curve and system curve. The pressure curve of the fan does not change. (Inlet and outlet dampers for centrifugal fans)

25

Figure 9. Fan operating points with outlet damper flow control method. (Inlet and outlet dampers for centrifugal fans)

2.4.4 Slip coupling

Slip coupling as a flow control method means any type of clutch which can be used to adjust torque from motor to fan. Slip coupling is usual only in big applications, power ranging over 150 kW. There are many solutions available for slip coupling control. A suitable clutch or coupling is selected according to the needed torque and rotational speed. Clutches that doesn’t have any contact surface use eddy currents or magnetic fields for transmission. Between clutches discs there can also be liquid to transmit power. When this kind of hydraulic clutch operates the clutch discs are pressed together by the use of spring tension causing pressure rise in liquid. The bigger the pressure is in the liquid the more power is transmitted through the clutch. Using slip coupling is a lossy way to control volume flow because the power generated by the motor isn’t transferred to the fan in its totality. (Torque Limiters and Slip Clutches Information on GlobalSpec, Seppänen, 143)

2.4.5 Voltage control

26

Fan’s rotational speed can be changed by changing the voltage input to motor. This is usually implemented by using a triac which is a semiconductor. Triac can conduct current to both directions. It consists of two parallel connected thyristors which have shared grid control. Controlling voltage with a triac is not fast because of its structure. (Karjalainen, 2)

2.4.6 Two speed motor

Two speed motor has two individual windings and it can be used in two different rotational speeds. Two speed motor as a flow control method is best suitable when the application the motor powers has two demand levels for power. It is cheaper to purchase and install than a variable speed drive. Two speed motors are available in variable torque, constant torque and constant horsepower versions depending on the application. These motors are about 60 % more expensive than single-speed motors. Big two speed motors are not very efficient at low loads. (Electric Ideas Clearinghouse, 1-3)

2.4.7 Cycling on/off

When flow control method is cycling (on/off) the fan and motor are running at the speed corresponding to the frequency of the input voltage or they are not operating. This flow control method is suitable for example in removal of gas. For example when sensor detects an overrun in limiting value fan starts and removes the gas from the space. It shuts down when the air in the space is gas free again. Continuous starting and stopping of the fan wears down the machinery.

2.4.8 Inlet box damper

Flow control method inlet box damper uses same kind of dampers as the control method called outlet damper but inlet dampers are installed before the fan. These dampers causes whirl in gas flow and due to this fan can not generate maximal pressure and volume flow. Fan’s power demand is reduced also due to inlet dampers and the result is similar to

27

reducing rotational speed in fan which has no flow control. (Inlet and outlet dampers for centrifugal fans) Inlet dampers’ effect on fan’s operating point is illustrated in figure 10. When dampers are closed the volume flow to the fan is decreased and so is the power demand. In operation point C in figure X the dampers are fully open and when operation point moves to points D and E as the dampers are closed new pressure and power demand curves are created. System curve remains the same. (Inlet and outlet dampers for centrifugal fans)

Figure 10. Centrifugal fan’s operating points with inlet damper. (Inlet and outlet dampers for centrifugal fans)

2.4.9 Inlet vanes for centrifugal fans

Inlet vanes are blades installed before the impeller of fan to direct the flow entering the fan. Vanes can be adjustable and when flow is decreased also fan’s power demand decreases. Inlet vanes are cheap and easy to install and almost maintenance free flow control method. This flow control method is used in applications which include relatively big pressures and flow velocities. Inlet vanes generate a swirl which is parallel to the direction impeller rotates. Disadvantages for inlet vanes are weakening efficiency in wide

28

control range. Also the fan’s noise level doesn’t decrease because the fan rotates at the same speed regardless of the adjustment. (Puhaltimet teollisuudessa ja LVI-laitoksissa, II 5-6 ; Inlet and outlet dampers for centrifugal fans)

2.4.10 Pitch angle adjustment in axial fans

Pitch angle adjustment for axial fan changes the outlet flow from the fan. Angles of the fan’s blades can be changed during operation so adjusting the flow doesn’t need the fan to be stopped. If the angle is too big the fan or duct can be damaged. If the angle is too small the fan can not transfer enough power to gas. Characteristic curves in appendixes F1, F2, F3 and F4 illustrates the effect pitch angle adjustment has on to the fan’s characteristic curve, efficiency and power demand.(Inlet and outlet dampers for centrifugal fan ; inlet vane control & dampers of any type ; Puhaltimet teollisuudessa ja LVI-laitoksissa, I 9)

3. CENTRIFUGAL PUMPS Pumps used for transferring liquids are categorized into dynamic pumps, displacement pumps and other pumps. Dynamic pumps use different types of impellers, which are rotating wheels with blades, to generate kinetic energy for liquid. Displacement pumps use for example a piston to move liquid from pump case to discharge pipe. Other types of pumps are for example electromagnetic pumps and buoyancy pumps. In this thesis only centrifugal pumps are studied since they are the most common type of pumps in industry and the only type the PumpSave –program can calculate energy savings to. The main characters of a pump can be discovered from the pump’s characteristic curves. These curves are generated in a particular rotational speed and diameter of the impeller. Curves state out pump’s head and efficiency in volume flow’s function. Pump’s head means the total head which includes the heads demanded by suction inlet and discharge of the pump. Characteristic curves for different types and sizes of centrifugal pumps can be found at appendixes G1 to G17.

29

The density of the liquid does not affect on the head but it affects on the increase of the pressure in the pump as equation 28 shows. (VLV-koneet luentomoniste 1, 81) Δp = ρgh

(28)

When designing a pumping system one main goal is to get the pumps working in their best efficiency as much of the running time as possible. Figure 11 shows the typical operating areas for different types of pumps.

Figure 11. Typical areas of operation for different pump types. (Luukkanen 2001, 45)

3.1 Structure and operation of centrifugal pump Centrifugal pumps are the most commonly used pump type in industry. About 80 % of process industry pumping needs is covered by centrifugal pumps. Main parts of a centrifugal pump are impeller, shaft and casing. Impeller has backward curved blades. Pump generates pressure to the liquid with the help of centrifugal force. Liquid discharges

30

out of the pumps case to a discharge pipe. One benefit of a centrifugal pump is its wide operational range. Volume flow can vary from 1 cubic meter per hour to over 10000 cubic meters per hour. Head can vary from less than a meter to about 3000 meters and the temperature of the liquid can be up to 400 °C. Large centrifugal pumps can have efficiency as high as 80 %. Figure 12 shows the main components of a centrifugal pump. (Luukkanen 2001, 44)

Figure 12. Structure of a centrifugal pump. (Viholainen 2007, 11)

3.2 Pump hydraulics Next chapters introduce the basic theory and equations of pumping. Main purpose of pumping is to transfer liquid from one place to another. Processes which need pumping this often means that the liquid is needed to be pumped upwards through a pipeline. Pumps function is to generate the needed pressure so this is possible.

3.2.1 Pump power demand and efficiency

Power input to liquid by pump is presented as equation 29 shows. Affecting factors to power are volume flow and density of the liquid as well as the total system head and gravitational constant.

31

P = qv ⋅ ρ ⋅g ⋅ H

(29)

where P

power input to liquid [W]

qv

volume flow [m³/s]

g

gravitational constant [m/s²]

H

total system head [m]

Pump receives power usually from a squirrel cage motor via shaft and transfers it forward to liquid. Yet pumping is a lossy process so pumps demand for power can be expressed with equation

Ppump =

qv ⋅ ρ ⋅g ⋅ H

η pump

(30)

where Ppump pump’s demand for power [W] η pump efficiency of the pump

(Wirzenius 1978, 47-50)

3.2.2 System head and net positive suction head

While pumping liquid receives addition to total elevation head which is called the total system head and marked as H. Total system head can be divided into static head and dynamic head. Static head is the invariable part of the total head and it consist of geodetic head and the head which is result of pressure difference between suction inlet and discharge side of the pump. Geodetic head means the difference between liquid levels between suction side and discharge side. Static head can be presented with equation.

H st = H geo +

where

p 2 − p1 ρg

(31)

32

H geo

geodetic head [m]

H st

static head [m]

p2

pressure at discharge outlet [Pa]

p1

pressure at suction inlet [Pa]

Dynamic head is the part of the total head which is consists of pressure losses in the piping and the volume flow of the liquid. Dynamic head can be presented with equation

H dyn =

v 22 − v12 + ∑ Hr 2g

H dyn

dynamic head [m]

v1

flow velocity at suction inlet [m/s]

v2

flow velocity at discharge outlet [m/s]

(32)

where

∑H

r

friction head [m]

Friction head means the part of total head which is due to friction. It can be presented with equation H r = kq 2v

(33)

where

k

constant due to liquids characteristics

Friction head can also be presented as equation 7 shows. In equation 34 parts of piping which causes losses are taken into account. These parts are for example valves, curves and connections. Also the roughness of the pipes coating is taken into account. ⎛ l H r = ⎜ λ + ∑ζ ⎝ d

2

⎞v ⎟ ⎠ 2g

(34)

33

where

λ

friction constant due to roughness of pipings coating

ζ

friction constant due to valves, curves and connections in the piping l

d

total length of piping [m]

diameter of piping [m]

Total head can be presented as equation 35 shows after all factors are taken into account.

H = H st + H dyn = H geo +

p 2 − p1 v 22 − v12 + + ∑ Hr ρg 2g

(35)

Net positive suction head or NPSH informs how much pressure must exist at pump’s suction inlet in addition to liquids steam pressure so that the pump wouldn’t cavitate. NPSH can be expressed with equation p1 + p b − p v v12 + − H r ,1 − H 1,geo NPSH = ρ ⋅g 2g

(36)

where pb

barometric pressure [Pa]

pv

steam pressure in liquid [Pa]

H r ,1

head due to losses at suction inlent [m]

H 1,geo static suction head [m]

H 1,geod is negative in equation 36 if the pump is placed below the tank water level and has positive suction. Figure 13 clarifies NPSH on suction lift. Cavitation means that pump’s performance collapses producing lower head and flow than expected when steam bubbles form in the liquid. Bubbles collapse instantly and cause shock waves to the impeller and damage it. Strong corrosion will occur if cavitation continues for a longer period. Strongly cavitating pump will fill up of steam and isn’t able to produce pressure. Cavitation can be noticed from a rattling sound and vibration. Using several impellers behind each other or

34

connecting pumps in series reduces the possibility of cavitation essentially. (Volk 2005, 89 ; VLV-koneet luentomoniste 1, 102-104)

Figure 13. NPSH. (Viholainen, 19)

3.2.3 Performance curves and affinity laws

Characteristic curves for pumps are usually presented in q v , H -coordinates where is drawn with q v , H -curve also q v , P , q v , η and NPSH –curves. With these curves pump’s operation can be observed when different variables change. Pump’s characteristic curves are created for specific rotational speed and diameter of impeller. q v , H -curve illustrates the relationship between volume flow and head, q v , P -curve illustrates the relationship between power demand and volume flow, q v ,η -curve illustrates the relationship between efficiency and volume flow and NPSH –curve illustrates the change in demand of minimum suction head when volume flow changes. When pump’s rotational speed changes moves the original operation point of the pump to a new characteristic curve provided by the new rotational speed. This happens through an affinity parable which goes via origin. Figure 14 illustrates pump’s characteristic curves at

35

rotational speeds 0,8n, n and 1,2n. Pump’s other features are unchanged in other words impellers diameter, blade angles and piping stays unchanged. (VLV luentomoniste 1, 96)

Figure 14. Effects of rotational speed change on to the operating point of the pump. (VLV-koneet luentomoniste 1, 96)

Affinity parable can be presented with equation 2

⎛q ⎞ H = H 0 ⎜⎜ v ⎟⎟ , ⎝ q v0 ⎠

(37)

where H0

original head [m]

q v0

original volume flow [m³/s]

Efficiency of the pump doesn’t change when moving via affinity parable. Next affinity laws have been able to be assigned for pumps. (VLV luentomoniste 1, 96 - 97)

36

The relationship between volume flow and rotational speed is: qv n = q v0 n 0

(38)

where n

new rotational speed [rpm]

n0

original rotational speed [rpm]

The relationship between head and rotational speed is:

H ⎛ n ⎞ =⎜ ⎟ H 0 ⎜⎝ n 0 ⎟⎠

2

(39)

The relationship between power and rotational speed is:

P ⎛ n ⎞ =⎜ ⎟ P0 ⎜⎝ n 0 ⎟⎠

3

(40)

3.3. Flow control methods in pumping In process the pumping demand changes in accordance with how the plant is operated so pumping must adjust to the needs of the whole process. When selecting flow control method for pump, three things must be paid attention to. First of all what kind of affect does the control have on the volume flow on wanted pressure level. Secondly attention must be paid on the affect on the efficiency of pumping and last is the change in power demand. When adjusting flow in pumping it is important to be able to operate the pumps at their best efficiency. Most commonly used flow control methods are throttling control, changing rotational speed of the pump (variable speed control), cycling and organizing a current from discharge to suction inlet. (Karttunen 2004, 252)

37

3.3.1 Variable speed control

It is possible to adjust the rotational speed of a pump while the motor rotates at a constant speed. These methods are for example a hydraulic clutch, hydraulic clutch and moment converter, inductive clutch or a mechanic variator which however is uncommon solution. Some electric motors are possible to use at different speeds but variable speed control has begun to establish more and more position at pump control in many processes. (Karttunen 2004, 252, VLV-luentomoniste 1, 99, Sarkomaa 1997, 6) Hydraulic clutch can transfer motors power to pump at rotational speeds from 0 % to 80 % of maximum speed. At speeds 80 % to 100 % of maximum the clutch is locked so that the motor is connected via planetary transmission directly to the pump. Inductive clutch is an electrical equivalent to hydraulic clutch. Other half of the inductive clutch contains a shortcircuit where currents induce if it moves regarding to magnetic field generated by a coil. In order to generate a magnetic field direct current must be used. Inductive clutches transmission ability depends on slipping and the magnitude of the current. Induced currents and the power they generate are dependent on the slipping of the clutch. (Sarkomaa 1997, 7) Using frequency converter in pumping isn’t a new thing but it has become more and more competitive choice for flow control in pumping because of advancement in technology and reduced investment costs. Changing the rotational speed of the motor the head or the volume flow generated by the pump can be changed. Frequency converter controls the electric motor, which shaft the pump’s impeller is connected to, and its rotation and at the same time collects data from pumping. Figure 15 illustrates the main components of a variable speed drive controlled pumping. (Volk 2005, 373)

38

Figure 15. Variable speed controlled pumping. (Viholainen, 31)

Pump’s operation point moves when the rotational speed of the pump changes. Operation point is located at the intersection of pump and system curves. Figure 14 shows how volume flow and head changes when rotational speed is changed. Main benefits of variable speed control executed by a frequency converter are energy saving, possibility to replace an existing flow control method with a frequency converter without replacing the whole system, wide range of control without massive energy losses, fine tuning possibility, lower loads on bearings and on pumps and the possibility to start and stop the pump carefully. Variable speed control is the best choice for flow control in cases where the flow of pumping has to be changed often. (Viholainen 2007, 33; Volk 2005, 374-375) Other benefits of variable speed is that the pump’s efficiency doesn’t reduce during and afnter adjustments especially if systems static head is small. With variable speed control there is no need for a loss causing throttling valve. (Viholainen 2007, 32)

3.3.2 Throttling control

In throttling control flow is controlled by a throttling valve which is installed to pipeline after the pump. It is not recommended to install throttling valve to a suction line because it would cause possibility to cavitation. Throttling control is very common flow control

39

method in pumping because it is cheap and easy to implement. Using throttling control though reduces efficiency in pumping. Pump produces bigger head when volume flow is reduced. This growth in head is wasted energy. Throttling control is justified when the pump is small and there is not much range for adjustments. It can also be used when the pump operates with too little pressure compared to normal operating point. Figure 16 illustrates the components of throttling controlled pumping system. (Karttunen 2004, 253, VLV-koneet luentomoniste 1, 99)

Figure 16. Throttling controlled pumping system. (Viholainen, 29)

3.3.3 On/off –control

Cycling, or on/off –control, means that the pump either runs or doesn’t run. This control method can be used for example in situation where a tanks liquid level must be in between certain limiting values. If the liquid level drops below a certain limiting value, the pump starts and fills the tank to another limiting value. When it reaches the upper limiting value pump stops. Starting and stopping the pump strains especially when the pump gets a full load instantly. This could be resolved with variable speed control because the pump could be started and stopped with slowly accelerating and slowing.

3.3.4 Other control methods

40

Liquids speeds tangential component can be changed by bending the blades of impeller. This control method doesn’t cause much of losses but it also has very limited adjusting possibilities. It also requires the pump to be stopped. It is a feasible control method for axial pumps and in cases where flow resistance of piping is small. Volume flow can be controlled also by blades which are installed just before the impeller. This control method is complicated and expensive and it is feasible for only half axial or axial pumps in cases where resistance from piping forms a significant part of the total head. (VLV-luentomoniste 1, 100, Pumpputekniikka, I 18) It is also possible to create a flow from pump’s discharge outlet to suction inlet but this also isn’t a efficient way to adjust flow. This method is mainly used with propeller pumps. In order to adjust a pump’s volume flow changes can be made to the impeller by changing its diameter or by replacing the impeller to a whole new one which is geometrically different than the original one. Of course it needs the pump to be stopped and it is feasible only then when pump’s productive values are needed to be changed substantially and permanently. (VLV-luentomoniste 1, 100, Pumpputekniikka, I 18)

4 SQUIRREL GAGE MOTOR Popularity of the squirrel gage motor is based on its reasonable price, small maintenance costs, working reliability and tolerance for tough environmental conditions. Squirrel gage motors are manufactured for power ranging from 50 W to 10 MW. Rotational speed is commonly 10, 12,5, 25 or 50 rounds per second. Rotational speed of a squirrel gage motor depends on terminal pair number, resistance of the rotor, magnitude input voltage and frequency and slipping of the load. Motor with two terminals running in 50 Hz electric network runs at the speed of 3000 rpm, motor with four terminals runs at the speed of 1500 rpm and motor with six terminals runs at the speed of 1000 rpm. Previously when unusual rotational speed has been required the choice has been a direct current motor but nowadays frequency converter is preferred. Squirrel gage motor with power more than 11 kW has efficiency from 90 % to 96 %. Smaller squirrel gage motors have efficiency below 90 %.

41

Squirrel gage motors are usually air cooled but big motors with power more than one megawatt can also be water cooled. (Sarkomaa 1997, 8)

4.1 Structure of squirrel gage motor Squirrel gage motor is simple and robust by its structure. The stator of a squirrel gage motor has three-phase winding where alternate current is led. Current generates a rotating magnetic field at the motor’s air gap and the magnetic field induces current and voltage to the rotor which makes it rotate. (Montonen 1985, 16) Squirrel gage motor is the most common type of electric motors in industrial use. Nowadays lots of electric motors are being manufactured to be used with frequency converter. Winding on the rotor of squirrel gage motor is done by pressure moulding or by copper bars. It is called short circuit winding. /1/, /3/

4.2 Motor losses Loss power of the motor consists of sine losses and harmonic losses. This is presented with equation Ploss = Psin loss + Pharmloss

(41)

where Ploss

motor loss [kW]

Psin loss

sine losses [kW]

Pharmloss

harmonic losses [kW]

Sine losses are formed from nominal loss of motor, loss caused by the rotational speed and loss caused by the load.

42

2

Psin loss = 0,2 ⋅ Plossnom

⎛ T ⎛ f ⎞ ⎟⎟ ⋅ Plossnom + 0,65 ⋅ ⎜⎜ + 0,15 ⋅ ⎜⎜ ⎝ Tnom ⎝ f nom ⎠

2

⎞ ⎟⎟ ⋅ Plossnom ⎠

(42)

where Plossnom

motor nominal loss [kW]

f

output frequency of the drive [1/s] nominal supply frequency [1/s]

f nom T

actual torque of the motor [Nm]

Tnom

nominal torque of the motor [Nm]

Torque produced by the motor can be solved with equation

P = T ⋅ 2 ⋅π ⋅

n 60

(43)

Harmonic loss can be presented with equation 43. It is not dependent on the load but is reversely dependent of the switching frequency of the frequency converter.

Pharmloss =

9 ⋅ Pnom f sw

(44)

where f sw

switching frequency of the drive [1/s]

Pnom

motor nominal power [kW]

In calculation for the switching frequency value 3000 Hz can be used. (Ruuskanen, 51-52)

5. FREQUENCY CONVERTER Frequency converter’s main function in industry is to adjust the rotation speed of ACmotor. AC means alternate current. Adjustment is based in changing the voltage and

43

frequency of the current fed into the motor. With frequency converter the acceleration speed and retardation speed of the motor can be changed to match the needs of the process. Benefits of frequency converter drive use are decrement of strain in the electric network which feeds the process and decrease in energy consumption in constant use. /1/, /2/ Frequency converters can be divided into direct frequency converters and frequency converters with intermediate circuit. Direct frequency converters can be divided into cyclone converters and matrix converters. Frequency converters with intermediate circuit can be divided into commutated frequency converters, frequency converters with current intermediate circuit and frequency converters with voltage intermediate circuit. Frequency adjustment for squirrel gage motor is commonly handled by frequency converter with intermediate circuit. Structure of this frequency converter is illustrated in figure 17. (Hirvonen 2007, 1-2)

Figure 17. Block diagram of a frequency converter. (Montonen 1985, 20)

Rectifier changes the alternate current from the network into direct current. Rectifier can be uncontrolled implemented with diode bridge or controlled thyristor rectifier. Inverter which creates the alternating voltage of wanted frequency is placed next to the motor. Function of the intermediate circuit is to filter the direct current provided by the rectifier and to separate the rectifier and the inverter from each other and operates as a harmonic filter. Other common components of frequency converters are network filter, cooling fan, processor and some power electronic components. (Montonen 1985, 20-21) Load commutated frequency converter contains two thyristor bridges which are linked with intermediate circuit with chokers. Other thyristor bridge is connected to the network and other to the electric motor. Bridge next to the network operates as a rectifier and the bridge next to the motor operates as an inverter. In frequency converters with current

44

intermediate circuit the diode bridge produces the direct current to the intermediate circuit. Voltage for the motor is controlled by modulating the pulse width because diode bridge is uncontrollable. In frequency converter with intermediate circuit the bridge next to the motor consists of commutation capacitors. They help to generate steadier torque than produced by load commutated frequency converter. (Hirvonen 2007, 7-8)

5.1 Drive losses Power loss of a frequency converter can be calculated with equation 45. Drive loss consists of no-load basic loss, speed dependent no-load loss and load-dependent loss. ⎛ f Pdriveloss = 0,35 ⋅ Pdrivelossnom + 0,1 ⋅ ⎜⎜ ⎝ f nom

⎞ ⎛ T ⎟⎟ ⋅ Pdrivelossnom + 0,55 ⋅ ⎜⎜ ⎠ ⎝ Tnom

⎞ ⎟⎟ ⋅ Pdrivelossnom (45) ⎠

where Pdriveloss

drive loss [kW]

Pdrivelossnom

drive nominal loss [kW]

(Ruuskanen, 53)

6. FANSAVE AND PUMPSAVE ENERGYSAVING TOOLS With FanSave and PumpSave energy saving tools energy consumption when using different flow control methods can be compared. Comparison is always performed between variable speed drive and existing flow control method. Programs are based on Microsoft Excel and contain also some Visual Basic code. Both programs are quite easy to use and they have features that make the usage easier like fitting the view of the main window for the screen and saving the performed calculation as a separate file. Programs also have buttons for copying the suitable frequency converter model to clipboard and printing the calculation in hand. Calculation can be performed in SI-units and in American units. Programs also include a default model for the profile of the yearly operation time.

45

Calculation of energy costs is common feature in both programs and is performed the same way. In both programs energy price, investment cost, interest rate and lifetime of the application are input as initial values. As results programs provide consumed energy in both flow control methods and possible energy savings as a subtraction. Programs are also able to calculate yearly carbon dioxide emissions caused by the application. Production of CO2 can be defined per kWh. Other economic results are payback period and present value of the investment. Calculation of payback period is based on cumulative cash flow and present value informs if the investment is profitable or unprofitable.

6.1 FanSave FanSave compares energy consumption between variable speed drive controlled fan and other flow control methods. In calculation program creates power demand curve in function of volume flow which helps the comparison between different flow control methods. FanSave includes axial and centrifugal fans. For centrifugal fan there is also three different blade curvatures to choose from. These are forward curved blades, backward curved blades and radial blades.

6.1.1 Initial values

Program needs as input values data from fan, transmission and motor. Also profile for yearly operating time is needed to input for calculation of total energy consumption. Existing flow control method and type of frequency converter must also be selected from drop-down menus. Needed fan data along with fan type and blade curvature are nominal volume flow, generated pressure difference and efficiency. FanSave assumes that fluid is air with the density of 1,2 kg/m³. It cannot be changed from the mainpage. If transmission is used its efficiency can be input in transmission data. Choices for existing flow control method are outlet damper, slip coupling, voltage control, two speed motor, cycling (on/off), inlet box damper, inlet vanes for a centrifugal fan and pitch angle for axial fan.

46

Initial values for the driving motor are input voltage, power and efficiency. Voltage can be chosen from 1-phase 115 or 230 V or 3-phase 230, 400, 460, 500 or 690 V. FanSave ads 10 % safety margin to the needed motor power which it calculates by using the fan data. Next program needs information about the annual operation time. Total operation time and percentual amount of time each flow rate is being used are needed for the calculation of total energy consumption. Last data about the appliances is what type of frequency converter is wanted to control the fan and the motor. Drop-down menu gives five options for frequency converter type. They are ABB component drive (ACS50), ABB component drive (ACS150), ABB general machinery drive (ACS350), ABB standard drive (ACS550) and ABB industrial drive (ACS800). Program presents suitable model for frequency converter when this selection is made.

6.1.2 Calculations

FanSave’s calculation is studied in this part of the thesis. All the equations and factors of the program uses are introduced. FanSave assumes that the medium is air so the calculation is observed based on that. Shaft power of the fan is calculated with equation 46. It contains term K p which is calculated with equation 47.

Paks =

K p ⋅ ρ ⋅ q v ⋅ Δp ⋅ 10 1,2 ⋅ η fan ⋅ η transmission

Kp= 1−

0,0035 ⋅ Δp ⋅ η fan p atm

(46)

(47)

In practice term K p gets values which are very close to 1 so it doesn’t have much relevance in what comes to results. Term K p would come up with values that have some effect to results if the fan’s produced pressure difference would be abnormally high. Equation 46 also contains term for fluid density at the numerator and value 1,2 at the denominator. When flowing fluid is air these components evert each other.

47

FanSave does not take into account any changes in fluids temperature during blowing. Therefore also the change in density is not correctly catered for. Density cannot be entered to the program as an initial value but it can be changed at the calculation sheet. Calculation could be added with possibility to take into account the temperature and density changes in fluid. Energy consumed at certain flow is written:

E i = Pi ⋅

t i ⋅ t total 100

(48)

Energy consumed during the yearly operating period is a sum of energies consumed at each flow rates.

E total =

nom .flow

∑E

i = 20%

(49)

i

Possible energy saving is calculated by subtracting the total energy consumed by variable speed drive from energy consumed by existing flow control method. E säästö,kWh = E säätö ,total − E vsd , total

(50)

Savings in currency is solved by multiplication of energy price and saved energy. E säästö,€ = E säästö,kWh ⋅ h sähkö

(51)

Specific fan power is calculated by FanSave with equation 52. Constant factors before variables are to correct the unit of the result. FanSave deals with one fan at a time so the SFP should be calculated for only one fan.

SFP =

0,01η fan

0,001 ⋅ Δp ⋅ 0,01η transmission ⋅ 0,01η motor

(52)

48

Power demand for most flow control methods in FanSave is calculated by using equation 53. For variable speed drive the efficiency of the frequency converter needs to be noticed also as equation 54 expresses.

Pi = xi ⋅

K p ⋅ ρ ⋅ q v ⋅ Δp ⋅ 1000

(53)

1,2 ⋅ η fan ⋅ η transmission ⋅ η motor

Pi , vsd = x i ⋅

K p ⋅ ρ ⋅ q v ⋅ Δp ⋅ 100000

(54)

1,2 ⋅ η fan ⋅ η transmission ⋅ η motor ⋅ η freq.conv.

Power demand for each flow control method at certain flow is dependent on a correcting factor xi which gets values depending on the flow rate. These factors are presented in Table 1. In table 1 for outlet damper the F-wheel stands for impeller with forward curved blades, B-wheel for backward curved blades and R-wheel for radial blades. Correcting factors are not dependent on the inintial values input to the program and they have been defined empirically according to FanSave manual. Table 1. Correcting factors for power demand. Outlet damper Centrifugal VSD Nom.

F-wheel

B-wheel

Rwheel

Axial

Slip

Voltage

coupling

control

On/off

Inlet

Inlet

Pitch

box

vanes

angle

damper

1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,0

1,00

1,000

1,000

90 %

0,740

0,840

0,960

0,960

1,140

0,733

0,846

0,9

0,84

0,763

0,710

80 %

0,515

0,715

0,910

0,910

1,170

0,626

0,703

0,8

0,75

0,636

0,508

70 %

0,370

0,615

0,855

0,855

1,125

0,505

0,563

0,7

0,67

0,559

0,356

60 %

0,240

0,525

0,790

0,790

0,958

0,386

0,440

0,6

0,62

0,483

0,246

50 %

0,155

0,450

0,723

0,723

1,000

0,297

0,331

0,5

0,58

0,436

0,169

40 %

0,095

0,395

0,655

0,655

1,125

0,220

0,223

0,4

0,54

0,394

0,127

30 %

0,055

0,380

0,570

0,570

1,150

0,163

0,156

0,3

0,50

0,360

0,085

20 %

0,035

0,380

0,490

0,490

1,154

0,146

0,094

0,2

0,47

0,339

0,059

flow

49

Factors for outlet damper flow control method with centrifugal fan are exactly the same whether the fan has backward curved blades or radial blades. Outlet damper is the only control method which is affected by the shape of the blades. Power demand curves for every fan case studied with FanSave are always the same shape because of these factors. These curves are illustrated in figure 18. Inlet vanes can only be used with centrifugal fan and pitch angle adjustment is a feature of axial fans.

1,4

Correcting factor

1,2 1 0,8 0,6 0,4 0,2 0 20 %

30 %

40 %

50 %

Variable speed drive Outlet damper R-wheel Voltage control Inlet vanes

60 % Flow

70 %

Outlet damper F-wheel Outlet damper axial Cycling (on/off) Pitch angle

80 %

90 %

Nom. flow

Outlet damper B-wheel Slip coupling Inlet box damper

Figure 18. Power demand curves.

Power demand curves created by FanSave for different flow control methods in figure 18 are quite well matching to corresponding curves in figure 7. So at least the shape of the curves is correct. FanSave has an option to study the power demand of the fan application if no flow control method is used. In this situation the power demand is calculated with equation 55.

Pnocontrol =

K p ⋅ ρ ⋅ q v ⋅ Δp ⋅ 1000 1,2 ⋅ η fan ⋅ η transmission ⋅ η motor

(55)

50

FanSave calculates power demand for two speed motor controlled fan by replacing the correcting factor xi with 0,25 when flow rate is 60 % of the maximum or lower. At higher flow rates the power demand is calculated according to equation 60 and the factor xi equals 1. This is studied more closely on the laboratory test section of the thesis. FanSave’s calculation for on/off flow control method assumes that flow rate 20 % of the maximum requires 20 % power of the maximum power and 30 % flow rate requires 30 % power of the maximum etc. FanSave’s calculation for on/off flow control method would be more accurate if it would be similar to PumpSave’s calculation which is dependent on the total quantity of pumped water. When controlling fan with on/off flow control method the fan either operates at full speed or doesn’t operate at all. With the profile of the operating period the total quantity of air blown could be solved. This would make possible to solve the time the fan needs to blow the total quantity when operating at full speed. Now the duty time could be solved with the nominal operating point of the fan and the total energy consumption could also be solved.

6.2 PumpSave With PumpSave pump applications energy consumption can be compared between variable speed drive control and few other flow control methods. Comparison is presented also as bar diagram. PumpSave is designed for only centrifugal fans.

6.2.1 Initial values

PumpSave needs data from the pump, piping, driving motor, possible throttling valve, exisiting flow control method and the annual operating time. Needed information about the piping are the density of the liquid and static head. Density is 1000 kg/m³ by default and static head determines the system curve in further calculations. Needed data of the pump are nominal volume flow, nominal head, maximum head and efficiency. These values can be found from the performance curve of the pump. If the piping has a throttling valve after the pump the addition to the total head can be input

51

where it says head over open throttling valve. It contains also a help box for more information. For the driving motor exactly the same information is needed as in FanSave. Also the profile and the total annual operating time are input the same way as in FanSave. For existing flow control method there is a drop-down box like in FanSave. For pump application the choices are throttling control, cycling (on/off) and hydraulic control. Frequency converter type can be selected as in FanSave and the options are the same.

6.2.2 Calculations

This part of the thesis introduces all the equations and factors PumpSave uses for its calculations. PumpSave solves pump’s power demand with equation 56. The program adds a 10 % safety margin for motor power.

Pp =

ρ ⋅ q v ⋅ H nom ⋅ g 3600 ⋅ η pump η transmission

(56)

Consumed energy for certain flow is calculated as equation 57 shows. Power required for certain flow is multiplied by the time pump was used in that flow

E i = Pi ⋅

t i ⋅ t total 100

(57)

Yearly energy consumption is calculated by adding all the energies used at all flow rates. This is presented in equation

E total =

nom .flow

∑E

i = 20%

i

(58)

Possible energy saving is calculated by subtracting the energy used by variable speed drive from the energy used by an existing flow control method. This is presented in equation

52

E säästö,kWh = E säätö,total − E vsd, total

(59)

Saving in currency is solved by multiplying the amount of saved energy by the price of energy as equation E säästö,€ = E säästö ,kWh ⋅ h sähkö

(60)

Payback period of the investment is calculated by means of cumulative cash flow. Investment cost is divided by the yearly saving as equation

t takai sin maksu =

I

(61)

E säästö,€

Throttling control

Power demand for throttling control for each flow rate is calculated as presented in equation 62. Change in head produced by pump is taken into account by the term H pi and the change in pump’s efficiency is taken into account by the term k pi . Calculation for power demand and these variables are presented in equations 62, 63 and 64.

Pthi ,i =

ρ ⋅ q v ,i ⋅ H pi ,i ⋅ g 3600 ⋅ k pi ,i ⋅ η pump ⋅ η motor

,

(62)

where H pi ,i = H max

k pi ,i

⎛q − ⎜⎜ v ,i ⎝ qv

2

⎞ ⎟⎟ ⋅ (H max − H nom ) and ⎠

1,44 ⋅ q v ,i ⎛ q v ,i ⋅ ⎜⎜ 2,4 − qv ⎝ = qv

⎞ ⎟⎟ ⎠

(63)

(64)

53

On/off flow control method

For on/off flow control method the most essential thing is the total amount of pumped water. By equation 65 can be calculated the amounts of pumped water at each flow rates.

Q i = q v ,i ⋅

ti % ⋅ t total 100

(65)

All the pumped amounts are next added together so the total amount which is pumped can be solved. This can be done with equation

q v , total =

nom.flow

∑Q

i = 20%

i

(66)

Next PumpSave calculates the amount of water that would be pumped if the pump would operate at full speed through the yearly operating time. This is presented in equation q v ,nom = q v ⋅ t total

(67)

Now can be calculated the percentage of the total operating time which pump has to operate at full speed to achieve the required amount of pumped water. This duty time can be solved with equation

t duty =

q v ,total q v ,nom

(68)

Next PumpSave calculates the time in hours which the pump has to operate at full speed. This is done with equation t on / off = t duty ⋅ t total

Finally the total energy consumed is calculated as equation

(69)

54

E on / off =

t on / off ⋅ Ppump

(70)

η motor

Hydraulic flow control

Power demand for each flow rate for hydraulic flow control is calculated by using equation 71. It contains term H si ,i which is the system head term and it is calculated by equation 72. Because clutch is used the power transmitted to pump’s shaft changes during flow adjustment. This is taken into account by term η transmission ,i which is calculated by equation 73. When pump operates at full flow the term η transmission ,i equals 0,98.

Phydr ,i =

ρq v,i ⋅ H si,i ⋅ g 3600 ⋅ η pump η motor ⋅ 1 ⋅ η transmission,i

H si ,i = H static

η transmission ,i

⎛ q v ,i + ⎜⎜ ⎝ qv

(71)

2

⎞ ⎟⎟ ⋅ (H no min al − H static − H overopenthrottlingvalve ) ⎠

⎛ ⎛ = 0,98 ⋅ ⎜ 0,2 20% flow ...0,9 90% flow + ⎜ 0,8 20% flow ...0,190% flow ⋅ 0,75 ⋅ ⎜ ⎜ ⎝ ⎝

(72)

⎛ H max ⎜⎜ ⎝ H nom

⎞ ⎞⎟ ⎞⎟ ⎟⎟ ⎟ ⎠ ⎟⎠ ⎠

(73)

Variable speed drive

Power demand for variable speed drive pump application is calculated by equation 74 for each flow rate. Term H si ,i in equation is the same factor for system curve as in hydraulic flow control method. Term k mi,i is correcting factor for driving motor’s efficiency. It is calculated with equation 75. Term k vsdi,i is correcting factor for frequency converter’s efficiency and it uses values shown in table 2.

55

Pvsdi,i =

k mi,i

ρ ⋅ q v,i ⋅ H si,i ⋅ g 3600 ⋅ η pump η motor η vsd ⋅ k mi,i ⋅ k vsdi,i

⎛q ⎞ = ⎜⎜ v,i ⎟⎟ ⎝ qv ⎠

0, 2−

0 , 2⋅H static H nom

(74)

(75)

Table 2. Correcting factors for the efficiency of the frequency converter.

Flow

Factor

Nom. flow

1,00

90 %

1,00

80 %

1,00

70 %

1,00

60 %

0,99

50 %

0,99

40 %

0,99

30 %

0,98

20 %

0,97

Energy consumption of variable speed drive adjusted pump for each flow rate is calculated with equation 76. Total energy consumption is addition of energy consumption of each flow rate. This can be calculated as equation 77.

E i , vsd = Pi , vsd ⋅

E vsd, tot =

t i ⋅ t total 100

(76)

nom.flow

∑E

i = 20%

i , vsd

(77)

7. COMPARISON BETWEEN FANSAVE’S CALCULATION AND PERFORMANCE CURVE BASED CALCULATION

56

This part of the thesis compares the calculations of FanSave energy saving tool to calculation based on different sizes and types of axial and centrifugal fans. Compared flow control methods are outlet damper and variable speed drive. FanSave does not take into account any changes in temperature or density of the fluid. Used performance curves for fans are created at constant density at and therefore changes in temperature or density were not taken into account. FanSave calculates the power demand as a total power demand for the whole application including driving motor and possible frequency converter. Changes of motor’s efficiency were taken into account by equations 41, 42, 43 and 44 and changes of frequency converter by equations 45 in characteristic curve based calculation. Operational data for each motor was found from website of motor manufacturer called Tamel. Outlet damper flow control method was observed with performance curves by choosing a nominal operating point from fan’s curve. Flow was then throttled 10 % at a time and new operating point was found via fan curve. Operational values of the nominal point were entered in FanSave as an initial values. Reading values from performance curves is not absolutely precise so the results may contain small errors on one way or the other. Characteristic curve based calculation gave good image about the accuracy of FanSave’s calculation and its applicability for different types of fans with different flow control methods. All the used performance curves for fans are found at appendixes E1 to E14 and F1 to F9.

7.1 Centrifugal fans Centrifugal fans for comparison were chosen from manufacturers called Fläktwoods and Hürner Funken which provided good performance curves for their fans. Some curves include correcting factors for fan’s efficiency when rotational speed changes. These factors were taken into account when the total power demand was calculated.

57

In calculation comparison the fans are divided based on the curvature of the blades. Four models with radial blades, six models with backward curved blades and four models with forward curved blades were chosen for the comparison.

7.1.1 Radial blades

Nominal point for each fan was chosen at operational area where the flow rate is as large as possible and the fan operates at good efficiency. Operation range was also pursued to keep as wide as possible.

HF R 560 – 13 D/R

First fan in this series is 560-13 D/R which is powered by 5,5 kW electric motor. It operates at the efficiency of 85 % at the nominal operating point Table 3. Calculations for HF R 560 – 13 D/R. Flow

Outlet Damper flow control n=1060 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

4,00

690

73 %

4,30

4,,29

3,70

760

1060

4,19

4,31

90 %

3,60

765

74 %

4,02

4,12

3,33

640

970

3,23

3,19

80 %

3,20

830

75 %

3,95

3,90

2,96

490

860

2,27

2,22

70 %

2,80

870

76,5 %

3,71

3,67

2,59

375

750

1,60

1,58

60 %

2,40

920

78 %

3,42

3,39

2,22

325

690

1,25

1,04

50 %

2,00

960

73 %

3,04

3,10

1,85

200

550

0,75

0,67

1,48

120

410

0,48

0,41

flow

40 %

0,79 0,78 0,77 0,76 0,75 0,74 0,73 0,72 0,71 0,7

5 4 3 2 1 0 40 %

50 %

60 %

70 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Efficiency

Power demand [kW]

58

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 19. Power demand and efficiency curves for HF R 560 – 13 D/R centrifugal fan.

Results in table 3 and in figure 19 show that for this fan FanSave’s calculation matches the characteristic curve based calculation extremely well for outlet damper flow control. Fan’s efficiency increases first when flow is choked for about 5 % and then decreases slowly towards the end of the operating range. Efficiency of the motor remains high during the adjustment. Results for power demand for fan operating under variable speed drive does not correlate as well as for fan with outlet damper. Power demand based on fan’s performance curve is a bit higher than power demand calculated by FanSave at all the operation points. Fan’s efficiency remains constant during the speed adjustment and motor’s efficiency decreases about 35 % during the adjustment. Drives efficiency decreases about 8 %. All in all FanSave’s calculation is very accurate for this fan case.

HF R 200 – 48 R

Second fan in this series is 200-48 R and it is powered by 22 kW motor. Motor operates at 88 % efficiency at the nominal operating point. Table 4. Calculations for HF R 200 -48 R. Flow

Outlet Damper flow control n=1640 rpm

VSD flow control

q_v

delta

[m³/s]

p [Pa]

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa]

59

Nom.

2,16

5400

70 %

18,45

18,69

1,72

5900

1640

15,58

15,30

90 %

1,94

5600

71 %

17,00

17,94

1,55

5000

1430

12,26

11,32

80 %

1,73

5850

74 %

15,44

17,01

1,38

3900

1230

8,87

7,88

70 %

1,51

6100

71 %

14,46

15,98

1,21

3000

1060

6,23

5,62

60 %

1,30

6400

66 %

14,01

14,76

1,03

2200

940

4,34

3,67

50 %

1,08

6450

62 %

12,59

13,51

0,86

1500

820

2,91

2,37

0,69

950

630

1,93

1,45

Power demand [kW]

40 %

20,00

0,75

15,00

0,7

10,00

0,65

5,00

0,6

0,00 40 %

50 %

60 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Efficiency

flow

0,55 70 % 80 % 90 % Nom. flow Flow Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 20. Power demand and efficiency curves.

For outlet damper the power demand curves are not as well matching in this case than in previous. Power demand calculated by FanSave is higher than power demand calculated with fan’s performance curve through the whole operating range. Fan’s efficiency increases first when the flow is throttled and it may be one reason for the difference in power demands. Motor’s efficiency decreases only about one percentage during the throttling. For variable speed drive controlled fan the results for power demand are perfectly matching. Fan’s efficiency decreases first slowly and more rapidly when flow rate enters below 70 %. Motor’s efficiency decreases about 39 % and drives about 8 %. For this fan case FanSave’s calculation is accurate for variable speed drive controlling and a bit inaccurate for outlet damper controlled flow.

HF R 710 – 13 D/R

60

Next fan under inspection is 710-13 D/R which is powered by 75 kW squirrel cage motor. It operates at the efficiency of 94,7 % at the nominal operating point.

Table 5. Results HF R 710 – 13 D/R. Flow

Outlet Damper flow control n=3000 rpm

VSD flow control eta_fan=58 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

13,00

2780

70 %

54,17

54,15

10,40

3400

1700

47,59

47,67

90 %

11,70

3120

75 %

51,11

51,99

9,36

2800

1530

35,73

35,27

80 %

10,40

3450

80 %

47,16

49,28

8,32

2100

1340

24,40

24,55

70 %

9,10

3680

80 %

44,08

46,30

7,28

1600

1170

16,85

17,50

60 %

7,80

3875

80 %

39,89

42,78

6,24

1220

1010

11,73

11,44

50 %

6,50

4050

77 %

36,21

39,15

5,20

850

830

7,51

7,39

60,00

0,82 0,8 0,78 0,76 0,74 0,72 0,7 0,68 0,66 0,64

50,00 40,00 30,00 20,00 10,00 0,00 50 %

60 %

70 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Flow

80 %

90 %

Efficiency

Power demand [kW]

flow

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 21. Power demand curves for HF R 710 – 13 D/R centrifugal fan.

For this fan the results for power demand in outlet damper flow control are quite well matching. Difference between power demands increases little bit when fan’s operation enters smaller flows. Efficiency of the fan increases first and then starts to decrease. Fan’s efficiency remains good through the whole operating range. Results for power demand when fan is operating under variable speed drive correlate extremely well. Efficiency of the fan decreases only few percentages during speed adjustment. Motor’s efficiency decreases about 13 % and drives efficiency about 7 %.

61

HF R 1000 – 13 D/R

Last fan equipped with radial blades is 1000-13D/R. It is connected to a 160 kW motor which operates at the efficiency of 95,9 % at the nominal point. Table 6. Calculations for HF R 1000 – 13 D/R. Flow

Outlet Damper flow control n=3000 rpm

VSD flow control eta_fan=58 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

24,50

3640

75 %

124,76

122,82

22,00

4000

1300

116,80

115,48

90 %

22,05

4000

81 %

114,25

117,91

19,80

3100

1150

82,42

85,46

80 %

19,60

4220

80 %

108,50

111,77

17,60

2400

1015

57,72

59,47

70 %

17,15

4450

79 %

101,41

105,01

15,40

1900

900

41,45

42,40

60 %

14,70

4680

78 %

92,67

97,03

13,20

1350

730

26,52

27,72

50 %

12,25

4780

77 %

80,08

88,80

11,00

960

620

17,00

17,90

0,82

140,00 120,00

0,8

100,00 80,00 60,00 40,00

0,78 0,76

Efficiency

Power demand [kW]

flow

0,74

20,00 0,00

0,72 50 %

60 %

70 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency

Flow

80 %

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 22. Power demand curves for HF R 1000 – 13 D/R centrifugal fan.

Results for power demand when fan operates under outlet damper flow control are quite well matching as figure 22 show. Power demand calculated by FanSave decreases slightly slower than power demand calculated with the fan’s performance curve. Fan’s efficiency behaves similarly as in previous cases. It increases first when the flow is throttled and then starts to decrease. Changes in efficiency depend on the chosen nominal operating point.

62

Good matching of the power demand curves point that this approach and this choice for nominal point match FanSave’s calculation. Results for the variable speed drive controlled fan are similar to previous cases. Power demand curves between calculations are almost identical. Fan’s efficiency remains high during the adjustment only dropping about a percentage. Motor’s efficiency decreases nearly 10 % and drives about 8 % during the speed adjustment.

7.1.2 Fans with forward curved blades.

Next series of fans under the lens are models with forward curved blades. Like in the section of fans with radial blades also in this series the study starts with the least powerful fan and advances to more powerful models.

GTLF-3-025

First fan with forward curves blades is GTLF-3-025 and its impellers diameter is 250 mm. Fan is powered by 4,8 kW motor which operates at 80 % of efficiency at the nominal operating point. Table 7. Calculations for GTLF-3-025. Flow

Outlet Damper flow control n=3000 rpm

VSD flow control eta_fan=58 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

0,91

2250

58 %

4,17

4,39

0,88

2250

3000

4,10

4,31

90 %

0,82

2240

57 %

3,81

3,69

0,79

1800

2700

3,01

3,19

80 %

0,73

2230

55 %

3,50

3,14

0,70

1420

2400

2,18

2,22

70 %

0,64

2220

53,5 %

3,15

2,70

0,61

1080

2100

1,53

1,58

60 %

0,55

2210

52 %

2,79

2,31

0,53

800

1800

1,06

1,03

50 %

0,46

2200

51 %

2,39

1,98

0,44

560

1500

0,72

0,67

40 %

0,36

2200

49 %

2,04

1,74

0,35

370

1200

0,50

0,41

30 %

0,27

2200

48 %

1,63

1,67

0,26

200

900

0,35

0,24

flow

0,8

5 4 3 2 1 0

0,6 0,4 0,2

Efficiency

Power demand [kW]

63

0 30 %

40 % 50 %

60 %

70 %

Flow

80 % 90 % Nom. flow

Calculated power demand for outlet damper

Power demand for outlet damper by FanSave

Calculated power demand for VSD Fan efficiency in outlet damper

Power demand for VSD by FanSave

Figure 23. Power demand and efficiency curves.

For VSD flow control the results for power demand are correlating through the whole operating range. A little difference is generated at the smallest and highest possible flows. Efficiency of the fan remains constant during the speed adjustment. Efficiency of the motor decreases about 55 % and the efficiency of the drive about 8 % during the adjustment. For outlet damper flow control the results for power demand are not as well matching as in vsd flow control. There is no big difference generated between the calculations but the power demand curves are in a bit different shape. Fan’s efficiency decreases about 10 % during the adjustment. Motor’s efficiency decreases nearly 8 %. However the calculation of FanSave for outlet damper flow control is quite well matching for this type of fan.

GTLF-3-040

Second studied fan is GTLF-3-040. It is a bit more powerful than the previous model and has impeller the size of 400 mm of diameter. Driving motor for the fan is 9 kW squirrel cage motor which reaches the efficiency of 87 % at the nominal operating speed. Table 8. Calculations for GTLF-3-040. Flow

Outlet Damper flow control n=1700 rpm

VSD flow control eta_fan=65 %

q_v

delta

[m³/s]

p [Pa]

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa]

64

Nom.

2,20

1800

64 %

6,89

7,08

1,95

1800

1700

6,17

6,31

90 %

1,98

1800

65 %

6,13

5,95

1,76

1500

1540

4,72

4,67

80 %

1,76

1800

64 %

5,56

5,07

1,56

1050

1310

3,07

3,25

70 %

1,54

1800

62,5 %

5,01

4,36

1,37

900

1200

2,39

2,32

60 %

1,32

1800

62 %

4,37

3,72

1,17

600

1010

1,52

1,51

50 %

1,10

1800

59,5 %

3,84

3,19

0,98

450

850

1,08

0,98

40 %

0,88

1800

59 %

3,17

2,80

0,78

270

660

0,69

0,60

30 %

0,66

1800

58 %

2,51

2,69

flow

8 7 6 5 4 3 2 1 0

0,66

0,62 0,6 0,58

Efficiency

Power demand [kW]

0,64

0,56 0,54 30 %

40 %

50 %

60 %

70 %

80 %

90 %

Nom. flow

Flow Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 24. Power demand and efficiency curves.

When flow is controlled by an outlet damper the results for power demand are similar to previous case. Power demand calculated by FanSave decreases first more rapidly than in calculation based on the performance curve of the fan. Decreasing slows down when the operating area enters smaller flow rates giving the performance curve parabolic shape. Power demand curve based on the performance curve is more linear. Efficiency of the fan decreases only about 6 % during the adjustment. Motor’s efficiency decreases about 8 %. Results for vsd controlled fan are matching almost perfectly and there are no big differences in power demands between calculations. Efficiency of the fan remains constant during the adjustment and motor’s efficiency decreases about 37 % when the rotating speed decreases. Drives efficiency decreases about 9 %.

GTLF-3-056

65

Next fan model is GTLF-3-056. Its impeller diameter is 560 mm and the it is powered by 22 kW motor. Motor’s efficiency at nominal operating point is 88 %. Table 9. Calculations for GTLF-3-056. Flow

Outlet Damper flow control n=1300 rpm

VSD flow control eta_fan=65 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

4,50

2000

65 %

15,71

15,66

4,50

2000

1300

16,08

15,98

90 %

4,05

2000

63,5 %

14,49

13,16

4,05

1580

1140

11,65

11,83

80 %

3,60

2000

63 %

13,01

11,20

3,60

1230

1025

8,29

8,23

70 %

3,15

1985

61,5 %

11,63

9,63

3,15

990

900

6,07

5,87

60 %

2,70

1975

60,5 %

10,16

8,22

2,70

720

770

4,08

3,84

50 %

2,25

1975

59,5 %

8,71

7,05

2,25

500

640

2,70

2,48

40 %

1,80

1960

58 %

7,24

6,19

1,80

330

520

1,80

1,52

flow

0,66

18 16 14 12 10 8 6 4 2 0

0,62 0,6 0,58

Efficiency

Power demand [kW]

0,64

0,56 0,54 40 %

50 %

60 %

70 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 25. Power demand and efficiency.

Power demand curves for outlet damper controlled flow have the same shape as in previous cases. Performance curve based power demand is higher than power demand calculated by FanSave at all the operation points. Efficiency of the fan decreases about 7 % during the flow adjustment. Motor’s efficiency remains quite good also decreasing only about 6 %.

66

For vsd controlled fan the results for power demand are once again almost identical between calculations. Motor’s efficiency decreases about 31 % and drives about 9 %.

GTLF-3-071

Last fan with forward curved blades is GTLF-3-071 and it is equipped with impeller the size of 710 mm of diameter. The fan is powered by 22 kW squirrel cage motor which reaches the efficiency of 91,7 % at the nominal operating point. Table 10. Calculations for GTLF-3-071. Flow

Outlet Damper flow control n=850 rpm

VSD flow control eta_fan=69 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

7,70

1600

68 %

19,88

19,68

7,25

1590

850

18,69

18,52

90 %

6,93

1550

68 %

17,31

16,53

6,53

1230

760

13,13

13,71

80 %

6,16

1500

67 %

15,11

14,07

5,80

1000

680

9,62

9,54

70 %

5,39

1450

66 %

13,00

12,11

5,08

770

595

6,66

6,80

60 %

4,62

1420

65 %

11,13

10,33

4,35

560

505

4,37

4,45

50 %

3,85

1400

62 %

9,64

8,86

3,63

400

425

2,83

2,87

40 %

3,08

1400

58,5 %

8,25

7,77

2,90

255

330

1,73

1,76

30 %

2,31

1400

55,7

6,48

7,48

25

0,7

20

0,65

15 0,6

10 5

0,55

0

0,5 30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Efficiency

Power demand [kW]

flow

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 26. Power demand and efficiency curves.

67

Results for outlet damper flow control are seemingly better matching for this fan than the previous ones. There is very little difference between power demand curves at all points of operating range. Efficiency of the fan decreases about 12 % during the flow adjustment. Motor’s efficiency remains high during the adjustment decreasing only about 4 %. When fan is operating under variable speed drive the results for power demand between calculations are identical. Fan’s efficiency remains constant, motor’s efficiency decreases about 23 % and the drives efficiency drops about 8 %.

7.1.3 Fans with backward curved blades

Last series of studied fans consists of fans with backward curved blades. First fan in the series is the least powerful model and the last is the most powerful. Performance curves of the fans are found at appendixes E9 to E14. Some curves have correcting factors for fans efficiency during rotation speed adjustment. These factors were taken into account when executing calculations.

GTLB-3-025

First fan of this series is GTLB-3-025. Diameter of the fan’s impeller is 250 mm and it is connected to a 2 kW motor which operates at the efficiency of 71 % at the nominal operating point. Table 11. Calculations for GTLB-3-025. Flow

Outlet Damper flow control n=4500 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

0,78

1320

70 %

2,07

2,07

0,59

1800

4500

1,99

2,00

90 %

0,70

1570

74 %

2,10

1,98

0,53

1470

4100

1,52

1,48

80 %

0,62

1780

75,7 %

2,07

1,88

0,47

1120

3600

1,09

1,03

70 %

0,55

1840

75,5 %

1,88

1,77

0,41

850

3130

0,80

0,73

flow

68

60 %

0,47

2000

74 %

1,79

1,63

0,35

630

2670

0,58

0,48

50 %

0,39

2070

73 %

1,58

1,49

0,30

450

2300

0,44

0,31

1,35

0,24

295

1850

0,33

0,19

0,8

2,5 2

0,75

1,5

0,7

1

0,65

0,5 0

Efficiency

Power demand [kW]

40 %

0,6 40 %

50 %

60 %

70 % Flow Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 27. Power demand and efficiency.

Power demand calculated by characteristic curve of the fan is a bit higher than power demand calculated by FanSave. Difference is not big and the curves are the same shape so FanSave’s calculation matches the curve based power demand quite well. Fan’s efficiency increases first when the flow is throttled and starts to decrease when flow rate reaches 70 % of the maximum. Motor’s efficiency decreases only one percentage during the adjustment. Results for fan operating under variable speed drive are matching better though little difference is formed when operating point enters lower flow rates. This could be caused by changes in fan’s efficiency which decreases about 10 % during the adjustment. FanSave may not take into account any changes in fan’s efficiency when it is controlled by a frequency converter. Motor’s efficiency drops about 38 % and drives about 3 %.

GTHB-3-040

Next model with backward curved blades is GTHB-3-040. Diameter of the impeller is 400 mm and it is powered by 8 kW motor. Efficiency of the motor is 81 % at the nominal operating point.

69

Table 12. Calculations for GTHB-3-040. Flow

Outlet Damper flow control n=3500 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

2,38

2100

75 %

7,69

8,18

1,85

2850

3500

7,64

8,14

90 %

2,14

2450

79 %

7,66

7,86

1,67

2375

3170

5,83

6,02

80 %

1,90

2750

80,5 %

7,51

7,45

1,48

1880

2830

4,23

4,19

70 %

1,67

3100

80 %

7,45

7,00

1,30

1450

2500

2,99

2,99

60 %

1,43

3250

78 %

6,88

6,46

1,11

1070

2170

2,05

1,95

50 %

1,19

3300

76 %

6,00

5,92

0,93

715

1750

1,34

1,26

40 %

0,74

330

1300

0,76

0,77

30 %

0,56

240

1000

0,60

0,45

0,82 0,8 0,78 0,76 0,74 0,72 0,7 0,68

Power demand [kW]

10 8 6 4 2 0 30 %

40 %

50 %

60 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Flow

70 %

80 %

90 %

Efficiency

flow

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 28. Power demand curves for GTHB-3-040.

Results for power demand when fan is controlled by outlet dampers are rather well matching. At highest volume flows power demand calculated by FanSave is a bit higher but at lower flow rates performance curve based power demand is higher. Nevertheless the differences are very small so the FanSave’s calculation is pretty accurate for this type of fan. Efficiency of the fan increases first quite rapidly and then starts to decrease when flow rate moves lower than 80 % of the maximum. Fan can not operate at lower flow rates than 50 % of the maximum flow rate.

70

Results for power demand when flow is controlled by a variable speed drive are correlating extremely well. No big differences in power demands between calculations are formed at any point of operation. Fan’s efficiency decreases 8 %, motor’s efficiency 53 % and drives efficiency about 9 % during the speed adjustment.

GTHB-3-056

Next examined fan is GTHB-3-056. Its impellers diameter is 560 mm and it is connected to a 17 kW motor. Motor operates at 86 % of efficiency at the nominal operating point. Table 13. Calculations for GTHB-3-056. Flow

Outlet Damper flow control n=1900 rpm

VSD flow control eta_fan=84 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

4,45

2140

75 %

14,76

14,68

3,45

3070

2500

15,34

15,19

90 %

4,01

2650

80,5 %

15,34

14,10

3,11

2400

2240

10,95

11,24

80 %

3,56

2980

82 %

15,04

13,36

2,76

1920

2000

7,97

7,82

70 %

3,12

3200

80,5 %

14,39

12,55

2,42

1420

1740

5,39

5,58

60 %

2,67

3350

79 %

13,14

11,6

2,07

1070

1500

3,73

3,65

50 %

1,73

770

1280

2,52

2,36

40 %

1,38

500

1050

1,64

1,44

20

0,84 0,82 0,8 0,78 0,76 0,74 0,72 0,7

15 10 5 0 40 %

50 %

60 %

70 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 29. Power demand curves for GTHB-3-056.

Efficiency

Power demand [kW]

flow

71

Results for power demand when fan operates under outlet damper fan control differ between calculations. Based on the performance curve of the fan power demand is higher through the whole operating range. This is caused by high pressure difference generated by fan when the flow is throttled. Though the efficiency increases at first the power demand also increases. In FanSave’s calculation there is no possibility of power demand rising when flow is decreasing. Motor’s efficiency remains constant during the adjustment. For variable speed drive controlled fan the results for power demand are almost identical. Fan’s efficiency decreases only few percentages, motor’s about 28 % and drives efficiency about 8 %.

GTHB-3-071

Next model in this series is GTHB-3-071. Diameter of the impeller is 710 mm and it is connected to a 30 kW motor which operates at the efficiency of 93,2 % at the nominal operating point. Table 14. Calculations for GTHB-3-071. Flow

Outlet Damper flow control n=1900 rpm

VSD flow control eta_fan=84 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

8,0

2000

75 %

22,80

22,77

5,9

2900

1900

22,19

22,11

90 %

7,2

2400

81 %

22,80

21,89

5,3

2350

1720

16,33

16,36

80 %

6,4

2700

82,5 %

22,40

20,72

4,7

1800

1540

11,32

11,39

70 %

5,6

2900

83 %

20,94

19,47

4,1

1400

1320

7,91

8,12

60 %

4,8

3100

80 %

19,91

17,99

3,5

1000

1130

5,10

5,31

50 %

4,0

3200

76 %

18,06

16,46

2,95

700

950

3,29

3,43

40 %

2,4

480

790

2,14

2,10

30 %

1,8

270

600

1,29

1,22

flow

0,86 0,84 0,82 0,8 0,78 0,76 0,74 0,72 0,7

Power demand [kW]

25 20 15 10 5 0 30 %

40 %

50 %

60 % 70 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Efficiency

72

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 30. Power demand and efficiency curves.

Like in previous case also in this case power demand for outlet damper controlled fan based on the characteristic curve is higher than power demand calculated by FanSave through the whole operating area. In this case also the power demand of the fan does not start to decrease immediately after the flow is throttled because of the high pressure difference generated by the fan. Efficiency reaches its highest point when flow rate is 70 % and then starts to decrease rapidly. Fan’s operating range reaches 50 % flow of the maximum flow. Results for variable speed drive controlled fan are seemingly better than fan with outlet dampers. Power demand curves between calculations match perfectly. Efficiency of the fan decreases two percents. Motor’s efficiency decreases about 37 % and drives about 17 %.

GTHB-3-100

Second to last fan in this series is GTHB-3-100. Diameter of the fan’s impeller is 1000 mm and it is powered by 55 kW motor which reaches 94,2 % efficiency at the nominal operating point. Table 15. Results for GTHB-3-100. Flow

q_v

Outlet Damper flow control n=1360 rpm

VSD flow control eta_fan=82,2 %

delta

q_v

eta_fan

P_total

P_FanSave

delta

n

P_total

P_FanSave

73

[m³/s]

p

[kW]

[kW]

[m³/s]

[Pa] Nom.

p

[rpm]

[kW]

[kW]

[Pa]

15

2100

75 %

44,70

44,34

11,9

2850

1360

44,90

44,33

90 %

13,5

2500

81 %

44,70

42,57

10,7

2350

1230

33,63

32,81

80 %

12

2800

82,5 %

44,17

40,35

9,5

1800

1075

23,27

22,83

70 %

10,5

3000

83 %

42,57

37,91

8,3

1400

950

16,24

16,28

flow

80 %

7,1

1020

810

10,62

10,64

50 %

76 %

6,0

730

690

6,94

6,87

4,8

480

550

4,27

4,21

40 %

Power demand [kW]

50

0,84 0,82 0,8 0,78 0,76 0,74 0,72 0,7

40 30 20 10 0 40 %

50 %

60 %

70 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 %

90 %

Efficiency

60 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in vsd

Figure 31. Power demand curves for GTHB-3-100.

For more powerful fans the results for outlet damper seem to be quite similar and this case is not an exception. Based on the performance curve the power demand is higher than power demand calculated by FanSave. Cause is once again the rising pressure difference when the flow is choked. Operating range of the fan is quite narrow reaching only flow rate of 70 % of the maximum. Driving motor’s efficiency remains high through the adjustment. Power demands between calculations for variable speed drive controlled fan are well correlating for this fan also. Fan’s efficiency remains high during the adjustment. Motor’s efficiency drops about 18,5 % and drives about 10 %.

GTLB-3-140

74

Last centrifugal fan with backward curved blades is GTLB-3-140 and its impellers diameter is 1400 mm. The fan is powered by 75 kW motor which operates at 94,3 % efficiency at the nominal operating point. Table 16. Results for GTLB-3-140. Flow

Outlet Damper flow control n=900 rpm

VSD flow control eta_fan=85 %

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

25,70

1320

70 %

51,57

51,23

17,70

2360

900

53,56

52,81

90 %

23,13

1760

79 %

54,82

49,18

15,93

1920

815

39,66

39,08

80 %

20,56

2120

82,5 %

56,21

46,62

14,16

1500

720

28,10

27,20

70 %

17,99

2370

85 %

53,37

43,80

12,39

1200

640

20,22

19,39

60 %

15,42

2500

84 %

48,86

40,47

10,62

830

450

12,71

12,67

50 %

12,85

2650

82,5 %

44,01

37,04

8,85

600

460

8,36

8,19

7,08

380

370

5,08

5,02

Power demand [kW]

40 %

60

1

50

0,8

40

0,6

30 0,4

20

Efficiency

flow

0,2

10 0

0 40 %

50 %

60 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

70 % Flow

80 %

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 32. Power demand and efficiency curves.

Power demand for outlet damper controlled fan based on the performance curve of the fan is a lot higher than power demand calculated by FanSave. This result is valid at all other flow rates than at the nominal flow. Efficiency of the fan increases when the flow is

75

throttled at first and starts to decrease when flow rate goes below 60 % of the maximum. Motor’s efficiency remains at 94 % through the adjustment. When fan is operating under a variable speed drive the power demands between calculations are correlating. Efficiency of the fan remains constant during the speed change. Motor’s efficiency decreases about 22 % and frequency converter’s efficiency decreases about 11 %.

Conclusions

For variable speed drive controlled fans the results from this part of the thesis show that FanSave’s calculation is accurate. There are no big differences in power demands generated between calculations for any of the fans. Surprisingly though FanSave uses same correction factors for power demand for fans with radial and backward curved blades the calculation is still accurate. Driving motor’s and frequency converter’s efficiency changes during speed change is taken into account accurately and there is no need to adjust FanSave’s calculation for variable speed drive controlled fan. FanSave’s calculation is a bit inaccurate for fans with radial blades when they are controlled with outlet dampers. For most of the fans FanSave calculates too high power demand through the operating range. Difference between FanSave’s calculation and performance curve based calculations are not big. Correcting factors which FanSave uses for calculating power demand could be adjusted to a little bit lower level. For more accurate adjustments more examinations with different types and sizes of fans should be executed. Results for fans with forward curved blades show that FanSave’s calculation gives a bit too low values for power demand at most parts of the operating range when fan is controlled by outlet dampers. Curves created from values of power demands are in a bit different shape. Curve created from FanSave’s calculation is more parabolic and curve created from power demand values based on fans performance curve is more linear. Based on these observations FanSave’s correcting factors for power demand at lowest and highest flow

76

rates are accurate and little adjustment should be made when flow rate is between 40 and 80 % of the maximum. For outlet damper controlled fans with backward curved blades FanSave’s calculation produces a bit too low values for power demand at the possible operation range for most observed fan cases. Difference between calculations seems to grow larger as the fan sizes get bigger. So maybe in this case should be different correcting factors be used for more powerful fans than smaller fans. More accurate results would also be achieved by adjusting the existing factors. No changes are needed to make for FanSave’s calculation for fan controlled by a variable speed drive. Calculation for power demand is perfectly balanced and accurate for each fan case and this conclusion is valid for each studied fan regardless of blade curvature.

7.2 Axial fans This part of the thesis consists of comparison of power demand for different sizes of axial fans to FanSave’s calculation. Power demand for each fan is calculated by using the fans performance curves. These can be found from appendixes F1 to F9. First studied flow control method is pitch angle adjustment and the fans under the lens come from manufacturer called Hürner Funken. Nominal operating point for each fan was chosen from operating area of high volume flow and pitch angle. Flow was decreased 10 % at a time by moving to a suitable pitch angle curve. Fans were also studied by choosing the nominal point from the area of best efficiency but these studies did not match FanSave’s calculations as well as above-mentioned method. Results are presented in tables and illustrated as power demand curves in figures. Other studied fans come from manufacturers called Applied Energy and Ebm-Papst. Applied Energy XLCA1000 fan was chosen for inspection in order to get information from operation of a highly powerful fan. Ebm-Papst fans are inspected under outlet damper and variable speed flow control methods. First fan of the series is the least powerful and the

77

next fan in line is more powerful than previous one. Operational values for motors were looked up from motor manufacturer Tamels website.

HF A 630

HF A 630 is the first studied fan and it is the smallest one of the Hürner Funken axial fans. It is driven by a 1,1 kW motor at the efficiency of 73 %. Table 17. Results for HF A 630 axial fan. q_v

deltap

[m³/s]

[Pa]

eta_fan

Pitch

P_motorloss

angle

[kW]

eta_m

P_total

P_FanSave

[kW]

[kW]

[°] Nom.

3,35

152

65 %

50

0,21

79 %

0,99

1,00

90 %

3,02

140

70 %

45

0,17

78 %

0,77

0,74

80 %

2,68

125

70 %

40

0,14

77 %

0,62

0,53

70 %

2,35

115

70 %

35

0,13

75 %

0,52

0,37

60 %

2,01

105

70 %

30

0,12

70 %

0,40

0,26

50 %

1,68

88

75 %

25

0,11

65 %

0,32

0,18

40 %

1,34

75

60 %

20

0,11

60 %

0,28

0,13

30 %

-

-

-

-

-

-

-

0,09

20

-

-

-

-

-

-

-

0,06

1,20 1,00 0,80 0,60 0,40 0,20 0,00

100 % 80 % 60 % 40 % 20 % 0% 40 %

50 %

60 %

70 % Flow

Calculated power demand Fan efficiency

80 %

90 %

Nom. flow

Power demand by FanSave Motor efficiency

Figure 33. Power demand and efficiency curves.

Efficiency

Power demand [kW]

flow

78

Power demand based on pump’s performance curve is higher through almost whole operating range. At high flow rates the power demand is well matching between calculations. Fan’s efficiency remains quite high through the operating range but motor’s efficiency decreases nearly 20 %. This contributes to the difference in power demands.

HF A 800

This fan is driven by a 4 kW electric motor which operates at the efficiency of 86,5 % at the nominal operating point. Table 18. Results for HF A 800 axial fan. q_v

deltap

[m³/s]

[Pa]

eta_fan

Pitch

P_motorloss

angle

[kW]

eta_m

P_total

P_FanSave

[kW]

[kW]

[°] Nom.

6,85

245

70 %

50

0,32

88 %

2,71

2,80

90 %

6,17

225

75 %

45

0,27

87 %

2,12

2,00

80 %

5,48

205

75 %

40

0,25

86 %

1,74

1,43

70 %

4,80

200

78 %

35

0,23

84 %

1,46

1,00

60 %

4,11

165

75 %

30

0,22

81 %

1,12

0,69

50 %

3,43

145

70 %

25

0,21

77 %

0,92

0,48

40 %

2,74

105

60 %

20

0,21

70 %

0,69

0,36

30 %

-

-

-

-

-

-

-

0,24

20

-

-

-

-

-

-

-

0,17

3,00 2,50 2,00

100 % 80 % 60 %

1,50 1,00 0,50

40 % 20 %

0,00

0% 40 %

50 %

60 %

70 % Flow

Calculated power demand Fan efficiency

80 %

90 %

Nom. flow

Power demand by FanSave Motor efficiency

Figure 34. Power demand and efficiency curves.

Efficiency

Power demand [kW]

flow

79

Results for this fan are quite similar compared to the previous one. Power demand based on the characteristic curve is higher through most part of the operating range. It seems like FanSave does not take into account any changes in motor’s efficiency, because if it remains constant during the adjustment the power demand curves correlate seemingly better.

HF A 1000

Last fan from Hürner Funken is more powerful than previous ones. It is powered by 11 kW squirrel cage motor which operates at the efficiency of 91 % at the nominal operating point. Table 19. Results for HF A 1000 axial fan. q_v

deltap

[m³/s]

[Pa]

eta_fan

Pitch

P_motorloss

angle

[kW]

eta_m

P_total

P_FanSave

[kW]

[kW]

[°] Nom.

12,90

390

70 %

50

0,65

92 %

7,84

8,10

90 %

11,61

365

75 %

45

0,55

91 %

6,20

5,76

80 %

10,32

330

75 %

40

0,49

90 %

5,03

4,12

70 %

9,03

310

75 %

35

0,45

89 %

4,19

2,89

60 %

7,75

280

75 %

30

0,42

87 %

3,32

2,00

50 %

6,45

225

70 %

25

0,40

84 %

2,48

1,37

40 %

5,16

190

65 %

20

0,39

79 %

1,90

1,03

30 %

-

-

-

-

-

-

-

0,69

20

-

-

-

-

-

-

-

0,48

flow

10,00 8,00 6,00 4,00 2,00 0,00

100,0 % 80,0 % 60,0 % 40,0 % 20,0 % 0,0 % 40 %

50 %

60 %

70 % Flow

Calculated power demand Fan efficiency

80 %

90 %

Efficiency

Power demand [kW]

80

Nom. flow

Power demand by FanSave Motor efficiency

Figure 35. Power demand and efficiency curves.

For this fan also the results for power demand are similar as with previous fans. Characteristic curve based power demand is higher than power demand calculated by FanSave through the whole operation range. Fan’s efficiency remains quite good through the adjustment dropping about 10 % when entering the lowest possible flow rates. Efficiency of the motor decreases about 13 % during the pitch angle adjustment.

Applied Energy XLCA1000

The last fan which operation was studied under pitch angle adjustment is Applied Energy XLCA1000 axial fan. Driving motor for this fan produces 30 kW and has the efficiency of 93,2 % at the nominal operating point. Operation of this fan was observed with different approaches and compared to results from FanSave’s calculation. First approach was to choose the nominal point at operating area where fan operates at relatively high volume flow and efficiency. Flow was decreased 10 % at a time and operating point was moved to a suitable pitch angle curve. Fan’s efficiency decreased from 73 % to 65 % during the adjustment. This approach did not match FanSave’s calculation at all. Power demand was a lot higher based on the fan’s performance curve than in results of FanSave through the whole operating range. Next approach was to choose the nominal operating point at the operating area where pump operates at the best possible efficiency. For this pump this means 60 % flow of the maximum possible flow. Flow was decreased 10 % at a time just like before. Every

81

operating point was chosen so that the pressure generation would be as low as possible and the efficiency as high as possible. Therefore the power demand would be as low as possible. Still a large difference is formed between the power demands of the calculations. Also the operating range of the fan was very narrow. Last approach matched the FanSave’s calculation the best. Now the nominal operating point of the fan was chosen at area of high volume flow and pitch angle. The efficiency of the fan is lower at first but when the flow is adjusted smaller it gets better. Results for this case are presented in table 20 and as power demand and efficiency curves in figure 36. Fan’s performance curve can be found in appendix F4. Table 20. Results for XLCA1000. q_v

deltap [Pa]

eta_fan

[m³/s] Nom.

Pitch

angle

eta_m

[°]

P_total

P_FanSave

[kW]

[kW]

19,43

800

67,5 %

32

93 %

24,65

25,60

90 %

17,49

785

70,5 %

28

93 %

20,86

18,18

80 %

15,54

707

73,3 %

24

93 %

16,14

13,01

70 %

13,60

580

73 %

20

92 %

11,79

9,12

60 %

11,66

488

74 %

18

90 %

8,58

6,30

50 %

9,72

560

73,5 %

14

89 %

8,29

4,33

40 %

7,77

544

72 %

10

87 %

6,73

3,25

30 %

5,83

582

68 %

8

86 %

5,83

2,18

30,00

100,0 %

25,00

80,0 %

20,00

60,0 %

15,00 40,0 %

10,00

Efficiency

Power demand [kW]

flow

20,0 %

5,00 0,00

0,0 % 30 %

40 %

50 %

60 %

Flow

70 %

80 %

90 %

Nom. flow

Calculated power demand for pitch angle adjustment

Power demand for pitch angle adjustment by FanSave

Fan efficiency

Motor efficiency

Figure 36. Power demand and efficiency curves.

82

Ebm-Papst fans

Next flow control methods observed are outlet damper and rotational speed control handled by a frequency converter. Fan models come from manufacturer called Ebm-Papst and they are relatively small powered and their operating ranges cannot match the smallest flow rates calculated by FanSave. Each fan’s performance curve is marked with operation points which were used in observing the operation of the fan and in calculation of power demand. Outlet damper flow control method was studied by moving along fan curve to a new operating point. Rotational speed adjustment was studied by tracing the new operating point via system curve. Curves for fans are found in appendixes F5 to F9.

3G400

First observed fan in this series is 3G400 which is driven by 0,5 kW motor. It operates at efficiency of 72 % at its nominal operating point. Table 21. Results for 3G400. Flow

Outlet Damper flow control n=1660 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

1,27

88

64 %

0,23

0,24

1,27

88

1660

0,24

0,25

90 %

1,14

125

62 %

0,30

0,28

1,14

70

1450

0,18

0,18

80 %

1,02

160

59 %

0,35

0,28

1,02

55

1300

0,14

0,13

70 %

0,89

180

55 %

0,37

0,27

0,89

42

1150

0,11

0,09

60 %

0,76

190

53 %

0,34

0,23

0,76

30

1000

0,08

0,06

flow

0,400 0,350 0,300 0,250 0,200 0,150 0,100 0,050 0,000

70 % 60 % 50 % 40 % 30 % 20 %

Efficiency

Power demand [kW]

83

10 % 0% 60 %

70 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 % 90 % Nom. flow Flow Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 37. Power demand and efficiency curves for 3G400.

For outlet damper flow control method there is clear difference in power demands as can be seen from figure 37. Based on pump’s performance curve the power demand is higher than power demand calculated by FanSave through the whole operating range. Figure 37 contains also the efficiency curve for the fan in outlet damper flow control. Efficiency decreases about 11 % during the flow adjustment. Efficiency in speed control remains constant and therefore it is not illustrated in figure 37. Power demands in variable speed control are well matching for this pump case. A little difference starts to build up when operation enters smaller flow rates. All in all it is clear that for this case FanSave’s calculation is more correct in variable speed control than in outlet damper control.

3G990

Next fan is 3G990. It is powered by 1,1 kW electric motor which operates at 78 % of efficiency at the nominal operating point. Table 22. Results for 3G990. Flow

Outlet Damper flow control n=650 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

650

0,56

0,56

[Pa] Nom.

7,64

40

[Pa] 72 %

0,54

0,54

7,64

40

84

flow 70

71 %

0,84

0,62

6,88

35

580

0,45

0,51

80 %

6,11

93

67 %

1,05

0,64

6,11

26

565

0,32

0,29

70 %

5,35

116

60 %

1,29

0,61

5,35

19

450

0,23

0,20

60 %

4,58

135

54 %

1,44

0,52

4,58

15

400

0,18

0,13

1,600 1,400 1,200 1,000 0,800 0,600 0,400 0,200 0,000

80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0% 60 %

70 %

80 % Flow

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

90 %

Efficiency

6,88

Power demand [kW]

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 38. Power demand and efficiency curves for 3G990.

Fan curve for this case is very steep and it causes the difference in power demands when fan operates under outlet damper flow control method. Because the fan curve is so steep the pressure difference generated by the fan increases rapidly. Also the fan’s efficiency decreases about 18 % which also contributes to the difference between power demands. For rotational speed control the power demands are quite well matching. Power demand based on the fan’s performance curve is slightly higher than the power demand based on FanSave. Efficiency in speed control decreases about 2 %.

3G800

Next fan is 3G800 and it is second to last in this series. It is driven by 1,8 kW motor at the efficiency of 77 %. Table 23. Results for 3G800. Flow

q_v

Outlet Damper flow control n=1030 rpm

VSD flow control

delta

q_v

eta_fan

P_total

P_FanSave

delta

n

P_total

P_FanSave

85

[m³/s]

p

[kW]

[kW]

[m³/s]

[Pa] Nom.

p

[rpm]

[kW]

[kW]

[Pa]

5,56

110

63 %

1,19

1,26

5,56

110

1030

1,22

1,29

90 %

5,00

149

60 %

1,51

1,44

5,00

90

940

0,92

0,95

80 %

4,44

182

58 %

1,69

1,48

4,44

70

810

0,67

0,66

70 %

3,89

205

50 %

1,93

1,42

3,89

54

720

0,49

0,47

60 %

3,33

220

45 %

1,97

1,21

3,33

40

630

0,36

0,31

flow

70 % 60 %

2,000

50 % 1,500

40 %

1,000

30 % 20 %

0,500

Efficiency

Power demand [kW]

2,500

10 %

0,000

0% 60 %

70 %

80 %

90 %

Nom. flow

Flow Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 39. Power demand curves for 3G800.

For outlet damper in this case the results are similar to previous cases. Based on the fan’s performance curve power demand is higher than based on FanSave’s calculation through the whole operating range. Power demand rises through the adjustment according to the characteristic curve but according to FanSave the power demand starts to drop when flow rate is decreased below 80 %. Efficiency decreases about 18 % during the flow adjustment. During speed adjustment the power demands are well matching between calculations. Efficiency of the fan decreases only about 2 % when adjusting the flow.

3G650

Second to last fan of the series is 3G650. It is driven by a squirrel cage motor which produces 3 kW at the efficiency of 78 %.

86

Table 24. Results for 3G650. Flow

Outlet Damper flow control n=1350 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

5,19

105

74 %

1,01

0,94

5,19

105

1350

1,04

0,96

90 %

4,67

180

71 %

1,50

1,08

4,67

85

1200

0,81

0,71

80 %

4,15

235

67 %

1,81

1,10

4,15

65

1050

0,67

0,50

70 %

3,63

280

63 %

1,99

1,06

60 %

3,11

325

58 %

2,15

0,90

80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0%

Power demand [kW]

2,500 2,000 1,500 1,000 0,500 0,000 60 %

70 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

80 % Flow

90 %

Efficiency

flow

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave Fan efficiency in VSD

Figure 40. Power demand and efficiency curves for 3G650.

For this fan in outlet damper the power demands between calculations differs even more than in previous fan case. Fan’s power demand increases at smaller flow rates due to higher pressure development and decrease of efficiency. Efficiency drops about 16 % during the flow adjustment. In variable speed control fan’s operating range is very narrow for this fan. Power demand is slightly higher when calculation is based on fan’s characteristic curve than when power demand is calculated by FanSave. This result is valid through the whole operating range. Efficiency decreases 10 % during speed adjustment.

3G710

87

Last fan of the series is 3G710 and it is powered by 2,2 kW motor. It has 77 % efficiency at the nominal operating point. Table 25. Results for 3G710. Flow

Outlet Damper flow control n=1230 rpm

VSD flow control

q_v

delta

[m³/s]

p

eta_fan

P_total

P_FanSave

q_v

delta

n

P_total

P_FanSave

[kW]

[kW]

[m³/s]

p

[rpm]

[kW]

[kW]

[Pa] Nom.

[Pa]

5,47

152

71 %

1,48

1,52

5,47

152

1230

1,52

1,55

90 %

4,93

204

67 %

1,90

1,73

4,93

123

1100

1,14

1,15

80 %

4,38

257

63 %

2,27

1,78

4,38

98

970

0,84

0,80

3,83

75

880

0,61

0,57

flow

70 %

72 % 70 %

2,000

68 % 1,500

66 %

1,000

64 % 62 %

0,500

Efficiency

Power demand [kW]

2,500

60 %

0,000

58 % 70 %

80 %

Calculated power demand for outlet damper Calculated power demand for VSD Fan efficiency in outlet damper

Flow

90 %

Nom. flow

Power demand for outlet damper by FanSave Power demand for VSD by FanSave

Figure 41. Power demand and efficiency curves for 3G710.

This fan has very limited operating range. For outlet damper results for power demand are similar to previous cases. Fan reaches only 80 % of the maximum flow as the lowest flow rate. Efficiency decreases about 8 % during the adjustment When fan operates under variable speed drive the results for power demand correlate between performance curve based calculations and FanSave’s calculations. Efficiency remains at 71 % through the adjustment.

Conclusions

88

Results for pitch angle adjustment are similar for each fan case. Power demand calculated by FanSave is smaller than power demand calculated with each fans performance curve. This result is valid for each fan when flow rate is below selected nominal flow. Behaviour of fan efficiency is also quite similar in all cases. It increases first when flow is decreased and then starts to decrease when the operating range gets near its end. Anyway the changes are not drastic and don’t have much effect on the power demand. Efficiency of the driving motor is calculated by using equations 40, 41, 42 and 43. It decreases from 7 to 19 % depending on the fan case. Changes in motor’s efficiency affect on the total power demand. It is hard to say what changes in fan applications operation FanSave’s constant factors for power demand covers. However these results show that the factors should be revised. A little higher values would produce more truthful results for pitch angle flow control method. When controlling flow with outlet dampers the pressure difference generated by fan increases rapidly when flow is choked. Based on the performance curves of inspected fans the power demand increased through the whole operating range in all cases but one. On the other hand the operating ranges for the fans were so narrow that the power demand didn’t have time to decrease until the operating range reached its limits. There was much less changes in power demands for fans according to FanSave’s calculations. Power demand in FanSave’s calculation increased at first a little bit when the flow was choked and then decreased when flow rate went below 80 % of the maximum. Efficiencies of the fans decreased quite a lot at each fan case which also increases the power demand. Constant factors FanSave uses should be corrected so that the increase of power demand is more rapid at first when the flow is choked and the turning point where the power demand starts to decrease should be moved to operating area of smaller flow rate. When fans are controlled by a variable speed drive the results for power demand between calculations are well matching for each fan case. Efficiencies of fans decrease only few percents during the adjustment and thereby don’t have much effect on power demands. Driving motors efficiencies decrease a lot more but factors of FanSave seem to take this into account because there is not much difference in power demands.

89

Based on these examinations FanSave’s calculation is very accurate for variable speed drive controlled axial fans and not so accurate for axial fans with outlet dampers or pitch angle adjustment. When the calculation of power demand is inaccurate also the results for total energy consumption of the application is incorrect. How much it differs depends on the length and profile of the operation time.

8. LABORATORY TEST RUNS WITH CENTRIFUGAL FAN

8.1 Laboratory equipment To compare calculation results of FanSave –energy saving tool, laboratory test runs were performed at Lappeenranta university of technology. Fan which was used was a Ventur GMT-100T centrifugal fan equipped with straight radial blade impeller. The manufacturer of the fan didn’t have any information about the efficiency of the fan and there were no efficiency curves created either. Frequency converter used in the tests was a Mitsubishi Freqrol Z200 and the electric motor used in the tests was a Tamel SKg 71-2B squirrel cage motor. Manufacturer states that motor’s efficiency is 75 %. This value was used in FanSave’s calculation for comparison. Fans impeller was installed directly to the axel of the motor so there were no transmission losses generated. Input voltage to the motor was 400 V at 50 Hz and the motor and the fan rotated at 2790 rpm. Duct consisted suction pipe and discharge pipe which had an orifice plate for pressure measurement and a throttling valve for adjusting the flow. Diameter of the hole in the orifice plate was 59,998 mm and the diameter of the piping was 108,14 mm.

90

Figure 42. Ventur GMT 100 T centrifugal fans characteristic curve and dimensions (Ventur).

Flow could be adjusted with the choke valve at the discharge pipe and by adjusting the rotational speed of the fan with the frequency converter. Also the inlet box damper flow control was simulated by choking the inlet of the fan. Two speed motor as a flow control method was also simulated by operating the fan at two different rotation speeds. Frequency converter was used to adjust the speed. Pressure difference transmitters were used to measure the total pressure difference generated by the fan and the pressure difference at the both sides of the orifice plate. Pressure difference transmitters were connected to the orifice plate as corner tapping. The discharge pipe contained a thermoelement so the rise in the temperature could be measured. Fluke hydra data logger was used to measure the temperature and the pressure differences. Input power to the electric motor was measured with Fluke power logger 1735 measuring device. Rotational speed was calculated with the frequency shown on the display of the frequency converter.

91

Figure 43. Laboratory equipment in the university laboratory.

8.2 Measurements and comparison of results to FanSave’s calculation Measurement report is shown in appendix H1 for full results of the tests. Volume flow in all the test runs was calculated by using iteration according to SFS ISO 5167 standard. Tables for discharge coefficient and expansibility factor can be found in the appendixes C and D. In order to be able to compare the laboratory test results and the results received from FanSave’s calculation the initial values for FanSave were the same values produced by the fan used in laboratory tests. These test results are comparable only for same size centrifugal fans. More tests should be made with various size of equipment to attain allinclusive information about FanSave’s calculation.

8.2.1 Outlet damper

Throttling valve was used for adjusting the flow in throttling control and simulating the outlet damper flow control method. Due to the structure of the valve the flow could only be adjusted to about 70 % of the maximum flow. Fan was operated at its full capacity at first and the flow was reduced 10 % at a time. Results of the throttling valve test can be red from table 26. Table 26. Calculations for outlet damper controlled fan. Flow

q_v [m³/s]

delta

p

eta_fan

P_motorloss

eta_m

P_total

P_FanSave

92

[Pa] Nom.

[kW]

[kW]

[kW]

0,0941

1377

47 %

0,072

79 %

0,349

0,345

90 %

0,0846

1372

45 %

0,070

79 %

0,330

0,331

80 %

0,0758

1418

43 %

0,068

79 %

0,318

0,314

70 %

0,0678

1451

40 %

0,068

78 %

0,313

0,295

0,36

0,48

0,34

0,46 0,44

0,32

0,42 0,3

0,4

0,28

Efficiency

Power demand [kW]

flow

0,38

0,26

0,36 70 %

80 %

Calculated power demand

Flow

90 %

Nom. flow

Power demand by FanSave

Fan efficiency

Figure 44. Power demand curves for outlet damper controlled fan.

Results in table 26 and in figure 44 show that calculation of FanSave is well correlating with results from laboratory test when flow rate is 80 % of the maximum or higher. Entering lower flow rates test applications power demand curve straightens and starts to differ from curve created by FanSave’s calculation. Main cause for this is the decreasing of fan efficiency which speeds up a bit. One cause may be found from the constant factors FanSave uses for both backward curved and radial blades equipped fans.

8.2.2 Variable speed control

Rotational speed and volume flow through the fan was adjusted by frequency converter. Flow was decreased 10 % at the time starting with nominal flow. Throttling valve at the discharge pipe was open fully during this test. Results of the tests are presented in table 27. Testissä käytetyn taajuusmuuttajan nimellisteho oli 2,2 kilowattia, joten se ei toiminut sille ominaisella toiminta-alueella, mistä selittyy sen heikohko hyötysuhde. Tämä huomioitiin

93

myös FanSave –ohjelman laskennassa, jossa taajuusmuuttajan hyötysuhteelle syötettiin puhaltimen nimellisvirtausta vastaava arvo 91 %. Table 27. Calculations for variable speed drive controlled fan. FanSave q_v

deltap

[m³/s]

[Pa]

n

P_motorloss

[rpm]

[kW]

0,0914

1377

47 %

2970

0,072

79 %

0,034

91 %

0,383

0,382

90 %

0,0847

1185

50 %

2673

0,061

77 %

0,033

89 %

0,295

0,283

80 %

0,0759

1061

55 %

2376

0,052

74 %

0,031

86 %

0,231

0,197

70 %

0,0659

919

62 %

2079

0,045

69 %

0,030

83 %

0,173

0,140

60 %

0,0565

788

67 %

1782

0,040

62 %

0,029

79 %

0,136

0,092

50 %

0,0471

656

66 %

1485

0,037

56 %

0,028

75 %

0,112

0,059

40 %

0,0380

525

46 %

1188

0,036

55 %

0,028

74 %

0,107

0,036

30 %

0,0287

398

25 %

891

0,038

55 %

0,028

75 %

0,112

0,021

20 %

0,0190

218

5%

594

0,075

51 %

0,031

83 %

0,184

0,013

Nom.

eta_fan

eta_m

P_driveloss

eta_drive

[kW]

P_total

P [kW]

[kW]

0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0

Power demand [kW]

0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

Calculated power demand

60 % Flow

70 %

80 %

90 %

Power demand by FanSave

Efficiency

flow

Nom. flow Fan efficiency

Figure 45. Power demand curves for variable speed drive controlled fan.

Results show that the generated factors for power demand for this fan follows the factors at the FanSave calculations well as the flow decreases to 60 % of the nominal flow. Flows smaller than 60 % of the nominal flow the power demand starts to differ compared to FanSave calculations. Power demand for laboratory test application is greater at 20 % and 30 % of the nominal flow than at 40 % of the nominal flow. This is because the efficiency of the fan collapses rapidly. Efficiency curve is also shown in figure 45. Power demand

94

curve was generated to clarify the results and there is a clear difference between the shapes of the power demand curves. When the power demand differs from laboratory test application the amount of used energy differs also. This affects on the amount of used energy during the operation time of the fan. It also has effect on the energy costs. How much the energy consumption and the costs differ depends on the size of the application and the operation profile during the operation time.

8.2.3 Two speed motor

Two speed motor as a flow control method was simulated by using the frequency converter and operating the fan and the electric motor at two different rotation speeds. FanSave assumes that when the flow rate is 60 % or less of the nominal flow the shaft power demand is 25 % of the maximum power demand. Power demand for two speed motor controlled fan can be examined by using table 27 which includes the results from variable speed drive controlled test. According to FanSave 60 % flow rate needs 25 % power of the maximum. So 60 % flow demands about 0,107 kW power according to table 27 when power demand increase caused by frequency converter efficiency is ignored. For this fan case 25 % power of maximum power equals 0,08 kW of power and it is not enough for 60 % flow. So for this case FanSave’s calculation is inaccurate.

8.2.4 Inlet box damper

Inlet box damper flow control method was simulated as choking the inlet of the fan. Results can be seen in table 28 and figure 45. Also the power demand curves were created based on laboratory tests and FanSave’s calculations.

95

As can be seen from table 28 and power demand curves in figure 46 there is considerable difference between power demand for laboratory equipment and a fan application calculated by FanSave. Table 28. Calculations for inlet damper controlled fan.

Flow

q_v

delta_p

P_motorloss eta_m

P_total

P_FanSave

[m³/s]

[Pa]

[kW]

[kW]

[kW]

0,0942

1377

47 %

0,072

79 %

0,349

0,345

90 %

0,0845

1185

39 %

0,070

79 %

0,329

0,290

80 %

0,0759

1065

33 %

0,068

79 %

0,317

0,259

70 %

0,0657

926

26 %

0,066

78 %

0,298

0,231

60 %

0,0564

795

20 %

0,064

77 %

0,284

0,214

50 %

0,0468

649

15 %

0,062

77 %

0,269

0,200

40 %

0,0377

536

10 %

0,061

76 %

0,259

0,186

30 %

0,0284

398

6%

0,060

76 %

0,249

0,172

20 %

0,0188

225

2%

0,059

75 %

0,235

0,162

Nom.

eta_fan

0,4 0,35 0,3 0,25 0,2 0,15 0,1 0,05 0

0,5 0,4 0,3 0,2

Efficiency

Power demand [kW]

flow

0,1 0 20 %

30 %

40 %

50 %

Calculated power demand

60 % Flow

70 %

80 %

90 %

Power demand by FanSave

Nom. flow Fan efficiency

Figure 46. Power demand curves for inlet damper controlled fan.

Power demand calculated by FanSave is smaller through the operation area. This being also the total energy consumption in operating time is smaller in FanSave calculation depending on the length of the operating time and on the size of the application. Efficiency curve is also shown in figure 46 and it collapses rapidly when flow is decreased.

96

8.3 Energy consumption examination in operation time Energy consumption between FanSave’s calculation and results from laboratory tests are compared in this part of the thesis. Annual operation time is chosen to be 5000 hours and profile for the operation time was generated with FanSave’s default button for all other flow control methods except for outlet damper control. Outlet damper control was created its own profile for operation time because due to limitations of equipment power demand for smallest flow rates was not possible to test. Results from energy consumption comparison is presented as bar diagrams in figure 47. Results show that there are no big differences between fan tested in laboratory and similar fan calculated with FanSave. Though over 100 kWh difference is generated between calculations for outlet damper and inlet damper control methods. For outlet damper energy consumption is higher based on fan tested in laboratory and for inlet damper control FanSave calculates higher energy consumption.

Energy consumption 2000 FanSave Lab

Lab FanSave

kWh

1500 1000

FanSave Lab

FanSave Lab

500 0

Outlet damper Variable [kWh] speed control

Inlet box damper [kWh]

Two-speed motor

FanSave

1617

636

1157

1048

Lab

1483

592

1317

990

Figure 47. Energy consumption comparison between FanSave and laboratory equipment.

9 LABORATORY TEST RUNS WITH AN AXIAL FAN

97

9.1 Laboratory equipment Axial fan used in laboratory tests was connected to the same piping as the centrifugal fan with the exception that the suction pipe was not connected to the fan. Fan sucked air from the surrounding space. Used fan is a compact appliance consisting of a fan impeller, an electric motor and a frequency converter with a remote control panel. Specific information about the machine was not available so data from similar appliances was applied in calculations. Data for the driving motor was used from Tamel Sg71-2A squirrel cage motor which produces 0,7 kW and operates at the efficiency of 70 % at the nominal operating point. Devices used in measuring were the same used for centrifugal fan. Examined flow control methods are variable speed control, outlet damper and inlet damper. Though dampers are seldom used with axial fans they were simulated because of the opportunity to do so.

9.2 Variable speed drive Efficiency of the frequency converter was assumed to be 90 % at the nominal operating point of the application. Efficiencies were calculated by using the power losses calculated with equations 41, 42, 43, 44 and 45. Results are presented as power demand and efficiency curves in figure 48.

Variable speed drive

0,4 Power demand [kW]

0,35 0,3 0,25 0,2 0,15 0,1 0,05 0

1,000 0,900 0,800 0,700 0,600 0,500 0,400 0,300 0,200 0,100 0,000

Efficiency

98

20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow Measured power demand Fan efficiency Drive efficiency

Power demand by FanSave Motor efficiency

Figure 48. Curves for VSD control.

Results for power demand in figure 47 show that measured power is little bit higher than power demand calculated by FanSave when flow rate is below 90 % of the maximum. However the difference is very small so FanSave’s calculation is quite accurate for this type of fan operating under variable speed drive. Calculated efficiency of the fan behaves quite oddly reaching its highest values when flow rate is 30 and 40 % of the maximum. Assumption of motor’s efficiency may have been too high because of the shape of the efficiency curve in figure 48. First it decreases rapidly when speed decelerated and later the decreasing is not so fast anymore. Drives efficiency decreases quite rectilinearly through the operating range.

9.3 Outlet damper Outlet damper flow control was simulated in laboratory by throttling the flow 10 % at a time at the outlet of the fan. Driving motor operated at full 3000 rpm speed trough the test run. Power loss caused by the frequency converter was not taken into account. Results are presented as power demand and efficiency curves in figure 49.

99

0,4

0,1

0,3

0,08 0,06

0,2

0,04

0,1

0,02

0

Efficiency

Power demand [kW]

Outlet damper

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow

Measured power demand

Power demand by FanSave

Appliance efficiency

Figure 49. Curves for outlet damper adjustment.

Power demand for the application was constant through the test. Shape of the power demand curve created by FanSave’s calculation is in totally different shape. Power demands are matching when flow rate is 50, 60 or 100 %. Otherwise FanSave calculates higher power demand for this type of fan application. Efficiency illustrated in figure 49 is for the whole appliance. It is the result of multiplication of motor’s assumed efficiency and fan’s calculated efficiency. It is decreasing because the power demand of the appliance remains constant and the flow rate decreases. Pressure difference of the fan does not increase much during the throttling.

9.4 Inlet damper Inlet box damper flow control was simulated by throttling the flow from the intake of the fan 10 % at a time. Driving motor operated at full speed through the test and the power loss caused by the frequency converter was not taken into account. Results are presented as power demand and efficiency curves in figure 50.

100

Inlet damper 0,100 0,080 0,060 0,040 0,020

Fan efficiency

Power demand [kW]

0,350 0,300 0,250 0,200 0,150 0,100 0,050 0,000

0,000 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow

Measured power demand

Power demand by FanSave

Appliance efficiency

Figure 50. Power demand and efficiency curves.

Measured power demand is seemingly higher through the whole operating range than power demand calculated by FanSave. Measured power demand does not decrease very much during the adjustment when power demand calculated by FanSave splits in half when fan is operating at the lowest flow rate. Efficiency illustrated in figure 50 is for the whole appliance. It is the result of multiplication of motor’s assumed efficiency and fan’s calculated efficiency.

Conclusions

For variable speed control the results for power demands are quite well correlating between measuring and calculation of FanSave. Because FanSave uses constant factors to calculate power demand at different flow rates it is impossible to say how the program assumes the behaviour of efficiencies for different devices. For outlet damper flow control method performed test run does not give very realistic results. Measured power demand is constant through the test run and curve created from FanSave’s calculation is wavy and achieves high points at two different flow rates. Increase in pressure difference is very little. This may be caused by the leaks in the pipeline.

101

Power demand results for inlet damper flow control are not matching at all between measured values and FanSave’s values. Measured power demand remains quite high during the adjustment and decreasing of power demand calculated by FanSave is parabolic. Changes in FanSave’s calculation based on these test runs are not necessary. FanSave’s calculation for variable speed drive is proven to be accurate and damper flow control with axial fans is extremely rare. A lot more test runs with different types and sizes of axial fans should be performed to achieve comparable results for FanSave’s calculations.

10 PERFORMANCE CURVE BASED CALCULATION COMPARISON TO PUMPSAVE’S CALCULATION In this part of the thesis PumpSave’s calculation is compared to different types and sizes of centrifugal pumps. Calculation comparison is based on each pump’s characteristic curve. Flow controls observed are throttling control and rotational speed control driven by a frequency converter. Hydraulic control was left aside because it is not as common flow control method than throttling or rotational speed control. Reading pump’s characteristic curve is not absolutely precise so the calculations done based on the characteristic curves may contain little errors. PumpSave calculates power demand for throttling control according to equation 62 where changes in head are noticed by term H pi ,i and changes in pump’s efficiency are noticed by the term k pi ,i . Term H pi ,i is affected by pump’s nominal flow, throttled flow and maximum and nominal head as equation 63 shows. Equation 64 shows that pump’s efficiency changes are dependent on pump’s nominal flow and throttled flow. Calculation performed with pump’s characteristic curve also examines how these factors change during throttling and changes are compared to PumpSave’s calculation.

10.1 VSD PumpSave calculates power demand for variable speed drive controlled pump application as equation 74 shows. It pays attention to changes efficiencies of frequency converter and

102

driving motor when flow rate changes. PumpSave uses constant factors when calculating change in frequency converter’s efficiency. These factors are presented in table 2. PumpSave uses equation 75 to calculate change in driving motor’s efficiency. It depends on flow rate, static head and pump’s nominal head. Static head affects also on system curve which PumpSave calculates by equation 72. Static head and system curve was created for each pump case separately. Operation range for each pump was attempted to keep as large as possible. Therefore also the comparison to PumpSave’s calculation could be done in wide operation range. Change in pump efficiency was observed with affinity parable. Comparison between characteristic curve based calculation and PumpSave’s calculation was executed by choosing wide range of different types of pumps. Pumps were chosen from manufacturer called Goulds Pumps which offer wide range of pumps for different purposes. Company’s website provides a useful tool for selecting pumps. Review of pumps starts from least powerful model and next pump in line is always more powerful than previous one. Operational values for driving motor’s were looked up from electric motor manufacturer Tamels website. For efficiency changes of drive motor equations 41, 42, 43 and 44 were used. Frequency converter’s efficiency change was calculated with equation 45. Pumps characteristic curves can be found at appendixes.

Goulds 3298

First studied pump is Goulds CDS 5309-1 model 3298 which operates at high rotational speed and reaches quite high head. Static head was chosen to be 15 m for this pump and system curve was generated according to it. Driving motor was selected to operate at 37 kW power and 93,3 % efficiency at the nominal operating point.

35

120

30

100

25

80

20 60

15

40

10

Head [m]

Power demand [kW]

103

20

5 0

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow

Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 51. Power demand and head curves.

From power demand curves at figure 51 can be seen that the power demand curves between calculations correlate almost perfectly together. Head curves in figure are equal to system curves of calculations. These curves match also extremely well. Pump does not reach the smallest flow rate calculated by PumpSave. Pump’s efficiency changes for better at first when flow is decreased and decreases when operating point enters smaller flow rates. However the changes are so small that they don’t have effect on total power demand.

1,2

Efficiency

1 0,8 0,6 0,4 0,2 0 20 %

30 %

40 %

50 %

60 % Flow

70 %

80 %

90 %

Nom. flow

Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Figure 52. Efficiency curves.

Figure 52 shows driving motor’s and frequency converter’s efficiencies as curves between calculations. Based on characteristic curve based calculations motor’s efficiency is better

104

through most part of the operation range. Efficiency starts to decrease more rapidly when flow rate reaches about 40 % of the maximum. Based on PumpSave frequency converter’s efficiency remains better than when efficiency is calculated using equation 45 through the whole operating range. Curve created for drives efficiency calculated by PumpSave is always the same shape due to the constant factors PumpSave uses.

GOUDLS IC/ICB/ICP/ICV

Next compared pump produces relatively small volume flow but high head. Static head was chosen to be 35 m. Motor driving the pump is 55 kW producing squirrel cage motor which operates at 94 % efficiency at nominal operating point.

250

60,000

200

50,000 40,000

150

30,000

100

Head [m]

Power demand [kW]

70,000

20,000 50

10,000 0,000

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 53. Power demand and head curves.

Results in figure 53 show that PumpSave’s calculation for power demand is well matching at most points of operating range. When entering the smallest possible flow rates power demand according to characteristic curve is slightly bigger than PumpSave’s calculated power demand. This is caused by changes in pump’s efficiency. It collapses when flow and rotation speed decreases enough. Pump’s efficiency curve is illustrated in figure 53. PumpSave’s calculation doesn’t take into account any changes in pump’s efficiency when changing rotation speed. Figure 52 also shows that head curves between calculations correlate almost perfectly.

105

1,200

Efficiency

1,000 0,800 0,600 0,400 0,200 0,000 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Pump efficiency

Figure 54. Efficiency curves.

Figure 54 illustrates efficiency changes among devices. In this case pump’s efficiency curve is also included because pump’s efficiency changes considerably during speed adjustment. This also has an affect on total power demand of the application. Driving motor’s efficiency is better based on characteristic curve based calculation than in PumpSave’s calculation through whole operating area. Situation is the opposite for frequency converter. These changes in efficiencies does not really have an effect on the total power demand.

Goulds 3638

Next pump produces big volume flow but lower head than previous models. Pump operates at high 85 % efficiency at its nominal operating point. Static head and starting point of system curve was chosen to be 5 meters. Driving motor for pump operates at 90 kW power and 95 % efficiency at its nominal operating point.

90

40

80

35

70

30

60

25

50

20

40

15

30 20

10

10

5

0

Head [m]

Power demand [kW]

106

0 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow

Nom. flow

Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 55. Power demand and head curves.

For this pump the results for power demand and head are well matching through whole operating range between calculations. Pump does not reach the smallest flow rate calculated by PumpSave. Pump’s efficiency decreases about 1,5 % during the speed adjustment so it does not really affect on the total power demand. Pump’s lowest possible operating point sets to about 40 % of the maximum flow.

1,2

Efficiency

1 0,8 0,6 0,4 0,2 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Figure 56. Efficiency curves.

Motor’s efficiency for this case is better through whole operating range when it is calculated with equations 41, 42, 43 and 44 than when it is calculated by PumpSave.

107

Frequency converter’s efficiency starts to decrease sooner and faster when it is calculated with equation 45 than when PumpSave calculates it. Even so the differences are quite small so they only have a small effect on the power demands.

Goulds 3628

Next inspected pump is Goulds 3628. Static head was chosen to be 10 m. Driving motor produces 132 kW at the efficiency of 95,8 %.

140

100 90 80

100

70 60 50 40 30

Head [m]

Power demand [kW]

120

80 60 40

20 10 0

20 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 57. Power demand and head curves.

Figure 57 illustrates that for this pump also the power demand and head curves correlate almost pefcetly. Pump’s operating range reaches about 40 % of the maximum flow.

108

1,2

PumpSave

1 0,8 0,6 0,4 0,2 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Figure 58. Efficiency curves.

Efficiency curves created for figure 58 are similar to previous cases. Motor’s efficiency is better through the operating range when it is calculated using the values from pump’s characteristic curve than when it is calculated by PumpSave. For frequency converter the situation is opposite. Pump’s efficiency decreases only about a percentage during the speed adjustment.

Goulds 5500

Next pump is again more powerful than previous one. This pump is Goulds 5500 and it is driven by a squirrel cage motor which produces 220 kW at 96,5 % efficiency. Static head for this case was chosen to be 17 meters.

250,000

120

200,000

100 80

150,000 60 100,000 40 50,000

Head [m]

Power demand [kW]

109

20

0,000

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow

Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 59. Power demand and head curves.

Power demand curves shown in figure 59 are well matching for most parts of the operating range. A little difference develops at the small flow rates because of the decrease in pump’s efficiency. PumpSave doesn’t calculate any changes for pump’s efficiency so the power demand calculated by PumpSave is smaller than power demand based on pump’s characteristic curve. Head curves are identical between calculations.

1,200

Efficiency

1,000 0,800 0,600 0,400 0,200 0,000 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow Motor efficiency Drive efficiency Pump efficiency

Motor efficiency by PumpSave Drive efficiency by PumpSave

Figure 60. Efficiency curves.

Efficiency curves for pump, motor and frequency converter are illustrated in figure 60. Motor’s efficiency based on the values from pump’s characteristic curve remains better through the whole operating range than when it is calculated by PumpSave. For frequency

110

converter the situation is yet again the opposite. Pump’s efficiency starts to decrease quickly when flow is decreased to about 40 % of the maximum flow.

Goulds 3635

Next studied pump is Goulds 3635. Driving motor for the pump produces 280 kW and has the efficiency of 97 % at the nominal operating point. Static head and the starting point of

300,000

120

250,000

100

200,000

80

150,000

60

100,000

40

50,000

20

0,000

Head [m]

Power demand [kW]

the system curve are 16 meters.

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow

Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 61. Power demand and head curves.

Power demand curves and head curves are yet again very well matching between calculations as can be seen from figure 61. Pump does not reach the smallest possible flow rate calculated by PumpSave.

111

1,200

Efficiency

1,000 0,800 0,600 0,400 0,200 0,000 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Motor efficiency Drive efficiency Pump efficiency

Motor efficiency by PumpSave Drive efficiency by PumpSave

Figure 62. Efficiency curves.

Efficiency curves for devices are illustrated in figure 62. No difference in results for motor’s and frequency converter’s efficiencies compared to previous cases. In this case the motor’s and frequency converter’s efficiencies based on the characteristic curve are very close to each other through the whole operating range. Pump’s efficiency decreases about 6 percentages during the flow adjustment.

Goulds 3652

Next pump is Goulds 3652 which can produce relatively large volume flow. Driving motor for this pump produces 320 kW and operates at 97 % efficiency at the nominal operating point. Static head is chosen to be 20 meters.

350,000

60

300,000

50

250,000

40

200,000 30 150,000 20

100,000

Head [m]

Power demand [kW]

112

10

50,000 0,000

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 63. Power demand and efficiency curves.

Figure 63 illustrates the power demand and head curves between calculations. Again the results between calculations are well matching. Pump’s operating range stops at about 50 % of the maximum flow.

1,200

Efficiency

1,000 0,800 0,600 0,400 0,200 0,000 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Figure 64. Efficiency curves.

Efficiency curves for driving motor and frequency converter are shown in figure 64. In this case the results are similar but differences between calculations are not as big as in some other previous cases. Pump’s efficiency decreases about 3 % during the flow adjustment.

Goulds CWX

113

Second to last pump under the lens is Goulds CWX. It is driven by a squirrel cage motor which produces 400 kW at the efficiency of 97 % at the nominal operating point. Static head is 30 meters.

400

120 100

300 80

250 200

60

150

40

Head [m]

Power demand [kW]

350

100 20

50 0

0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow

Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 65. Power demand and head curves.

Results as power demand and head curves are shown in figure 65. Power demand calculated by pump’s characteristic curve is slightly smaller than power demand calculated by PumpSave at the middle part of the operating range. Reason for this can be found from the changes in efficiencies because the head curves are almost identical. Efficiency curves are illustrated in figure 66.

1,2

Efficiency

1 0,8 0,6 0,4 0,2 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow

Motor efficiency Drive efficiency Pump Efficiency

Motor efficiency by PumpSave Drive efficiency by PumpSave

Figure 66. Efficiency curves.

114

Results for motor’s and frequency converter’s efficiencies are similar to previous cases. Based on the characteristic curve motor’s efficiency remains better and frequency converter’s efficiency remains worse through the whole operating range compared to efficiencies calculated by PumpSave. Pump’s efficiency increases first when rotation speed decreases but starts to decrease when flow rate reaches about 60 % of the maximum flow.

Goulds 5500

Last observed pump is the most powerful pump of the series. It can produce the volume flow of over 2800 m³/h at the best efficiency point. Static head was chosen to be 7,62 meters and driving motor produces 900 kW and operates at 97 % efficiency at the nominal

900 800 700 600 500 400 300 200 100 0

90 80 70 60 50 40 30 20 10 0

Head [m]

Power demand [kW]

operating point.

20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow flow Calculated pow er demand

Pow er demand by PumpSave

Head

Head by PumpSave

Figure 67. Power demand and head curves.

Like previous cases the power demand and head correlate almost perfectly between calculations as can be seen from figure 67. Pump does not reach the smallest possible flow calculated by PumpSave.

115

1,2

Efficiency

1 0,8 0,6 0,4 0,2 0 20 %

30 %

40 %

50 %

60 % Flow

70 %

80 %

90 %

Nom. flow

Motor efficiency

Motor efficiency by PumpSave

Drive efficiency

Drive efficiency by PumpSave

Figure 68. Efficiency curves.

Results for efficiencies are also similar compared to previous cases. Motor’s efficiency is better through the operating range when it is calculated with the values from pump’s characteristic curve than when it is calculated by PumpSave. For frequency converter the situation is the opposite. Pump’s efficiency remains good during the adjustment only decreasing about a percentage.

10.2 Throttling Throttling control is examined next as a flow control method for centrifugal pumps. Pumps under examination are from manufacturer called Afton pumps Inc. First pump has the lowest power demand and the last one is the most powerful. This study is based on the characteristic curves of the pumps. All pumps operate at the rotational speed of 3000 rpm. When throttling the flow the new operation point of pump is found along the pump’s characteristic curve. Pump’s power demand can be solved when flow, head and efficiency is known. These factors at each operation point are compared to PumpSave’s calculation. Results are shown as power demand, head and efficiency curves. Static head is not taken notice and no imaginary system curves were created in characteristic curves. Noticing static head does not affect on the calculation of PumpSave when examining throttling control. PumpSave does not calculate any changes in driving

116

motor’s efficiency during throttling control. Changes in motor efficiency are taken notice of in calculations based on pump’s characteristic curve.

Afton ILVS 1.5x2-7A

First pump under inspection is Afton ILVS 1.5x2-7A. Diameter of the pump’s impeller is 184 mm. Driving motor produces 4 kW of power and its efficiency at nominal operating

4

52

3,5

50

3

48

2,5

46

2 44

1,5 1

42

0,5

40

0

38 20 %

30 %

40 %

50 %

60 % Flow

Calculated power demand Head

70 %

80 %

90 %

Head [m]

Power demand [kW]

point is 85 %.

Nom. flow

Power demand by PumpSave Head by PumpSave

Figure 69. Power demand and head curves.

Power demand and head curves for this pump are shown in figure 69. As can be seen from the power demand curves there is difference in the values between calculations. According to PumpSave the power demand decreases more rapidly than in calculations based on the characteristic curve of the pump when the flow is choked. When flow is 40 to 80 % of the maximum flow calculated power demand based on the characteristic curve is larger than the power demand calculated by PumpSave. Head curves are also illustrated in figure 69 and there is quite similar difference as in power demand. Head based on the pump’s characteristic curve is larger at the whole operating range. This has an affect on the power demand.

117

0,6

Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

Pump efficiency

60 % Flow

70 %

80 %

90 %

Nom. flow

Pump efficiency by PumpSave

Figure 70. Efficiency curves.

Efficiency curves based on PumpSave’s calculation and pump’s characteristic curve are illustrated in figure 70. According to PumpSave the efficiency remains higher during the whole throttling than pump’s characteristic curve indicates. PumpSave does not take into account the change in motor’s efficiency in throttling control. When using equations 41, 42, 43 and 44 to resolve power loss and efficiency of the motor it drops about 2 % during the throttling. Pump does not reach the smallest possible flow calculated by PumpSave.

Afton ILVS 1.5x2-9H

Next pump is a bit more powerful than previous one producing larger volume flow and higher head. It’s impellers diameter is 229 mm and driving motor operates at the power of

75

8 7 6 5 4 3 2 1 0

70 65 60 55 50 20 %

30 %

40 %

50 %

60 % Flow

Calculated power demand Head

70 %

80 %

90 %

Nom. flow

Power demand by PumpSave Head by PumpSave

Head [m]

Power demand [kW]

7,5 kW and 87,5 % efficiency rate.

118

Figure 71. Power demand and head curves.

For this case power demand curve based on PumpSave’s calculation is very linearly descending. Power demand curve based on pump’s characteristic curve on the other hand differs dramatically from PumpSave’s curve when flow rate is 40 to 60 % of maximum. This is caused by rapid collapse of efficiency at that operation area. Efficiency curves by PumpSave and pump’s characteristic curve are created in figure 72. Characteristic curve based power demand is higher than power demand by PumpSave through almost whole operating area. This is supported also by change in pump’s produced head during throttling. Head based on pump’s characteristic curve is also higher than head calculated by PumpSave at all points of operation range. However the head curves are more congruent than power demand curves.

0,6

Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow Pump efficiency

Nom. flow

Pump efficiency by FanSave

Figure 72. Efficiency curves.

Efficiency for this pump is clearly worse than efficiency calculated by PumpSave through whole operating range. When flow is throttled from 60 % to 50 % of maximum the efficiency collapses rapidly and causes a rise in power demand curve seen before. Efficiency of the motor drops about 1,5 % during the throttling.

Afton ILVS 1.5x3-11H

119

Next pump under inspection is Afton ILVS 1.5x3-11H and its impellers diameter is 279 mm. Applicable motor for the pump is a squirrel cage motor which produces 30 kW power

25

118 116 114 112 110 108 106 104 102

20 15 10 5 0 20 %

30 %

40 %

50 %

60 % Flow Calculated power demand Head

70 %

80 %

Head [m]

Power demand [kW]

and has an efficiency of 92,9 % at the nominal operating point.

90 %

Nom. flow Power demand by PumpSave Head by PumpSave

Figure 73. Power demand and head curves.

For this pump as can be seen from figure 73 the power demand calculated by pump’s characteristic curve is higher than PumpSave’s power demand during the whole operating range. The main factor causing this difference is the head generated by the pump. Pump’s characteristic curve is steeply descending when moving in area of larger flows. When the flow is choked to about 70 % of the maximum flow the pump’s characteristic curve is almost horizontal and the head does not increase much. PumpSave’s curve is more linear and the biggest difference in head is formed when flow is 70 % of maximum. The head by pump’s characteristic curve is about 2 meters higher than the head calculated by PumpSave.

120

0,7 0,6 Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Nom. flow

Flow Pump efficiency

Pump efficiency by PumpSave

Figure 74. Efficiency curves.

Efficiency curves between pump’s characteristic curve and PumpSave’s are illustrated in figure 74. Again there is clear difference how the efficiency changes during the throttling adjustment. Pump’s efficiency starts to decrease much quicker according to characteristic curve than PumpSave’s calculation. This contributes also to the difference in power demands. Drive motor’s efficiency decreases only about one percent during the throttling. Smallest flow the pump reaches is about 40 % of the maximum flow.

Afton ILVS 2x3-13H

Next pump is Afton ILVS 2x3-13H and it produces quite high head and a little bit higher volume flow than the previous pump. Impellers diameter is 343 mm and the driving motor operates at 55 kW power and efficiency of 94 %.

121

175 170

50

165 40

160

30

155 150

20

Head [m]

Power demand [kW]

60

145 10

140

0

135 20 %

30 %

40 %

50 %

60 % Flow

70 %

80 %

90 %

Nom. flow

Calculated power demand

Power demand by PumpSave

Head

Head by PumpSave

Figure 75. Power demand and head curves.

Results in figure 75 shows that pump’s characteristic curve which is shown as head curve in figure X is steeper than the head curve calculated by PumpSave. This causes the power demand to be higher than the power demand calculated by PumpSave at the range the pump is able to operate.

0,6

Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

Pump efficiency

60 % Flow

70 %

80 %

90 %

Nom. flow

Pump efficiency by PumpSave

Figure 76. Efficiency curves.

For this pump case the efficiency curves are quite congruent at the range the pump operates. Efficiency of the motor connected to the pump decreases about one percent during the throttling.

122

Afton ILVS 3x4-10

Next pump is Afton ILVS 3x4-10 with the impeller diameter of 254 mm. Motor connected to the pump is a squirrel cage motor and it produces power 55 kW at the efficiency of 94 %.

95

35

90

30 25

85

20 80

15 10

Head [m]

Power demand [kW]

40

75

5 0

70 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow

Nom. flow

Calculated power demand

Power demand by PumpSave

Head

Head by PumpSave

Figure 77. Power demand and head curves.

Differences in power demand and in head between PumpSave’s calculation and characteristic curve based calculation are similar to previous pump cases. At the most points of operation the head and therefore the power demand are higher based on characteristic curve than based on PumpSave’s calculation. Pump’s power demand descends below PumpSave’s calculated value at the smallest possible flow. Reason for this is shown in figure 78 where efficiency curves are illustrated.

123

0,8 0,7

Efficiency

0,6 0,5 0,4 0,3 0,2 0,1 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Pump efficiency

Pump efficiency by FanSave

Figure 78. Efficiency curves.

Efficiency curves between PumpSave’s calculation and characteristic curve are quite congruent for this pump case. When the flow is throttled to about 45 % of maximum the efficiency curves cross and for smaller flows the efficiency based on the characteristic curve is better than the efficiency calculated by PumpSave. This shows up also in power demand curves at flow rate of 40 % in figure 77 where power demand based on characteristic curve is lower than power demand calculated by PumpSave

Afton ILVS 4x6-11H

Next pump under inspection is Afton ILVS 4x6-11H. Diameter of the impeller is 279 mm. Driving motor operates at 90 kW power and at the efficiency of 95 %.

124

90

110 105

70 60

100

50

95

40 30 20

90

Head [m]

Power demand [kW]

80

85

10 0

80 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Nom. flow

Flow Calculated pow er demand Head

Pow er demand by PumpSave Head by PumpSave

Figure 79. Power demand and head curves.

Power demand and head curves are illustrated in figure 79. Once again the power demand and head curves based on the pump’s characteristic curve go through higher values than curves based on PumpSave’s calculation. Efficiency curves are illustrated in figure 80. Difference between head and efficiency curves are the reason the power demand based on characteristic curve differs from the power demand calculated by PumpSave.

0,8 0,7

Efficiency

0,6 0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

Pump efficiency

60 % Flow

70 %

80 %

90 %

Nom. flow

Pump efficiency by PumpSave

Figure 80. Efficiency curves.

Shapes of efficiency curves are similar to previous pump cases. Efficiency based on characteristic curve starts to decrease sooner than efficiency based on PumpSave’s calculation when flow is throttled. This contributes to the difference in power demands.

125

Driving motor’s efficiency stays high during throttling decreasing only few decimals of percent.

Afton ILVS 3x4-13L

Next observed pump is Afton ILVS 3x4-13L with the impeller of 343 mm of diameter.

90 80

180 160

70 60 50 40

140 120 100 80

30 20 10 0

60 40 20 0 20 %

30 %

40 %

50 %

Calculated pow er demand Head

60 % Flow

70 %

80 %

90 %

Head [m]

Power demand [kW]

Pump’s driving motor produces 110 kW at the efficiency of 95,8 %.

Nom. flow

Pow er demand by PumpSave Head by PumpSave

Figure 81. Power demand and head curves.

Head curves for this pump are almost perfectly matching between pump’s characteristic curve and calculation of PumpSave as can be seen from figure 81. However the power demands between these two are not so well matching. Power demand based on calculations done by characteristic curve is higher than power demand calculated by PumpSave at almost whole operating area. Only at the highest and lowest possible flow the power demands are matching. Difference is caused by difference in efficiencies between pump’s characteristic curve and PumpSave’s calculations.

126

0,7 0,6 Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. flow Flow Pump efficiency

Pump efficiency by PumpSave

Figure 82. Efficiency curves.

Figure X illustrates the efficiency curves created by pump’s characteristic curve and PumpSave’s calculations. Efficiency indicated by pump curve decreases more quickly than efficiency calculated by PumpSave. Drop in efficiency is not so rapid when pump is operating at small flow rates and characteristic curve based efficiency is better than efficiency calculated by PumpSave at the smallest possible flow. Driving motor’s efficiency decreases less than a percent during throttling.

Afton ILVS 4x6-13H

Last studied pump is Afton ILVS 4x6-13H and it is the most powerful pump of the series. Pump’s impeller has the diameter of 343 mm and it is driven by a squirrel cage motor which produces 132 kW of power at the efficiency of 95,8 %.

140

170

120

165

100

160

80

155

60

150

40

145

20

140

0

Head [m]

Power demand [kW]

127

135 20 %

30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow Calculated pow er demand Head

Nom. flow

Pow er demand by PumpSave Head by PumpSave

Figure 83. Power demand and head curves.

Results between calculations are not surprising at this point. Values for power demand and head based on pump’s characteristic curve are once again higher than PumpSave’s calculated values. Difference in power demands is affected by difference in efficiencies which is shown in figure 84.

0,8 0,7

Efficiency

0,6 0,5 0,4 0,3 0,2 0,1 0 20 %

30 %

40 %

50 %

Pump efficiency

60 % Flow

70 %

80 %

90 %

Nom. flow

Pump efficiency by PumpSave

Figure 84. Efficiency curves.

Effciency curves based on pump’s characteristic curve and PumpSave’s calculations are quite congruent as figure 84 shows. Still the efficiency calculated by PumpSave is better than efficiency based on characteristic curve at the whole operating range of the pump. Driving motor’s efficiency remains high through the throttling.

128

Conclusions

PumpSave’s calculations for throttling control are fairly competent based on these examinations. In throttling control PumpSave starts the throttling at the best efficiency point of operation and the efficiency does not get better at any point of throttling. So it is important to check with the pump’s characteristic curve what the point of nominal flow for the pump is where the throttling of the flow is planned to start. PumpSave’s calculation presents a bit too low values for pump’s characteristic curve through the whole operating area. Each pump’s characteristic curve is shown in figures as a head curve. For one pump the curves between PumpSave’s calculation and the pump’s characteristic curve are almost perfectly matching but for other cases the head values differ more or less. This also contributes to the calculated power demand which was higher based on calculations done by pump’s characteristic curve than calculated power demand by PumpSave for each pump through nearly whole operation range. Results for pump applications power demand presented in previous chapter is supported by results acquired from efficiency examination. Comparison of efficiency for each pump case showed that PumpSave’s calculation produces higher values than pump’s characteristic curve indicates. Each pump’s efficiency starts to decrease much sooner according to pump’s characteristic curve than in PumpSave’s calculation when throttling the flow. Based on acquired results PumpSave’s calculated values for efficiency and head should be corrected a bit. Curves created by means of these values should be steeper. Pump’s characteristic curve should rise more rapidly and efficiency curve should decrease more rapidly when throttling the flow. For more exact results much more pump cases should be studied so the calculation would be more universal for all possible pumps. For one of the studied pumps its efficiency collapses so much at one stage of throttling that its power demand at that point rose higher than at previous operating point where the flow rate was larger. This is of course a special scenario and PumpSave can not predict it but

129

these kinds of situations can happen. All in all PumpSave gives fairly good prediction for pump applications total power demand and energy consumption.

11 LABORATORY TEST RUNS WITH A PUMP In this part of the thesis PumpSave’s calculation is compared to results gathered from laboratory testing. Flow control methods studied were throttling control and rotational speed control by frequency converter. Tests were done in Lappeenranta university of technology where is a specific laboratory for testing pumps. Pump used in test was made by G.A. Serlachius Oy model DC 80/260. It was powered by Strömberg HXUR 328A2 B3 electric motor which produces 15 kW. Speed of the motor was controlled with ACS 800 variable speed drive. A layout of the laboratory equipment is illustrated in figure 85.

Figure 85. Laboratory equipment. (Viholainen, 60)

A venturi tube was used to measure the speed and volume of the flow. Pressures at pump inlet and outlet were measured to make possible to calculate the total head of the pump. Measured data was lead to Fluke Hydra Data Logger. Also the torque provided by the motor was measured with separate device. Rotational speed, input frequency and shaft power could be red from the panel of the variable speed drive.

130

11.1 VSD First flow control method under observation is variable speed drive. Pump was first driven at full speed and adjustments were done by decreasing the rotational speed so that flow produced by the pump was decreased 10 % at a time. Rotational speed was adjusted by regulating the input frequency at the panel of the frequency converter. All the valves in the pipeline were fully open during the speed adjustment. Results and the most important variables are presented in table 29. Power demand and head between calculations are illustrated as curves in figure 86. Pump’s efficiency curve is illustrated in figure 87. Table 29. Results of test run for VSD flow control. PumpSave q_v

p_imu

p_paine

deltap_vent.

Vääntö

n

f

H

eta_p

P_displ

P_calc

P

H[m]

[m³/s]

[Pa]

[Pa]

[Pa]

[Nm]

[rpm]

[Hz]

[m]

[%]

[kW]

[kW]

[kW]

0,0354

82500

177000

50413

60,0

1500

50

9,63

35,5

9,60

9,43

9,41

9,63

0,0319

88125

171000

40700

47,5

1311

43,7

8,45

40,5

6,48

6,52

7,39

8,28

0,0283

92500

162000

32098

38,0

1158

38,6

7,08

42,7

4,55

4,61

5,71

7,06

0,0248

96250

156000

24975

30,6

1035

34,5

6,09

44,7

3,26

3,32

4,32

5,99

0,0212

99375

150000

18130

24,2

915

30,5

5,16

46,3

2,29

2,32

3,24

5,07

0,0177

102500

145500

12580

18,8

801

26,7

4,38

48,3

1,55

1,58

2,34

4,28

0,0142

105000

141000

8140

14,5

699

23,3

3,67

48,2

1,04

1,06

1,65

3,64

0,0106

106875

138000

4625

11,2

609

20,3

3,17

46,4

0,69

0,71

1,12

3,14

0,0071

108125

136500

2035

8,5

537

17,9

2,89

42,1

0,47

0,48

0,71

2,79

12

10

10

Power demand [kW]

12

8

8

6

6

4

4

2

2

0

Nostokorkeus [m]

131

0 20 %

30 %

40 %

50 %

60 % Flow

70 %

80 %

90 %

Nom. Flow

Akseliteho tm:n näytöltä

Laskettu akseliteho

Akseliteho by PumpSave

PumpSave akseliteho hyötysuhteen ka:n mukaan

Nostokorkeus

Nostokorkeus by PumpSave

Figure 86. Shaft power curves and head curves.

From curves illustrating shaft power in figure X can be seen that the shaft power red from the display of the VSD correlates with the calculated power demand. Calculated power demand is calculated on the basis of the measured values. Apposed to that shaft power calculated by PumpSave differs from measured and calculated shaft power through almost whole operating range. This is caused by changes in pump’s efficiency which PumpSave does not take into account. PumpSave is fed with the efficiency of the pump at the nominal operating point and it remains constant at each point of operation. There is also a power demand curve in figure X which illustrates power demand calculated by PumpSave and it is calculated with an average efficiency of the operation range. This curve matches the curves created with results from laboratory tests much better than curve created with efficiency at the nominal operating point. With this curve the biggest difference to curves from laboratory tests is formed when pump operates at maximum flow. This is because pump’s efficiency is worse than average in actual situation. Figure X shows also the head produced by the pump and head calculated by PumpSave. Values are well matching through the operation range though the head values calculated by means of the pressure difference of the pump are a bit higher at most part of the operation range. However the difference is so small that it hardly affects the total power demand of the application.

132

0,6 0,5

Efficiency

0,4 0,3 0,2 0,1 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow

Figure 87. Pump efficiency curve.

Figure 87 illustrates pump’s efficiency curve during the speed adjustment. Efficiency reaches its maximum value when the flow rate is 50 % of the maximum. At the nominal flow pump’s efficiency is at its lowest being only 35,5 %. This takes PumpSave’s calculation on the wrong track. Efficiency is not constant when pump’s rotation speed is changed and this causes difference on results for power demand.. Next is examined situation where point of nominal flow is also the best efficiency point. Nominal point is chosen to be 50 % of the maximum flow based on the previous study. Results of power demand and pump efficiency are illustrated in figure 88.

1,8

0,5 0,48

1,4 1,2

0,46

1

0,44

0,8 0,6

0,42

0,4

0,4

0,2 0

0,38 40 %

50 %

Akseliteho näytöltä Pump efficiency

60 %

70 % Flow

80 %

Laskettu akseliteho Poly. (Pump efficiency)

90 %

Nom. flow

Akseliteho by PumpSave

Efficiency

Power demand [kW]

1,6

133

Figure 88. Shaft power and efficiency curves.

Results show that all the power demands are now well matching. This way of inspecting pump’s operation matches best PumpSave’s calculation. This also was proved in section of the work where PumpSave’s calculation was compared to calculation based on pump’s characteristic curve. Figure X also shows points and trend line of pump’s efficiency.

11.2 Throttling Other flow control method which could be studied at the laboratory was throttling control. Pump operated at full speed through the test and at the nominal point all the valves on the pipeline were fully open. Flow was adjusted by throttling the last valve before the upper tank so that flow was decreased 10 % at a time. Results of calculations are shown in table 30 and as shaft power and head curves in figure 89. Efficiency curves of the pump are illustrated in figure 90. Table 30. Results for throttling control. PumpSave q_v

p_imu

p_paine

deltap_vent.

Vääntö

H

eta_p

P_displ

P_calc

P

H[m]

eta_p

[m³/s]

[Pa]

[Pa]

[Pa]

[Nm]

[m]

[%]

[kW]

[kW]

[kW]

0,0354

82500

177000

50413

60,0

9,63

35,5

9,51

9,43

9,41

9,63

35,5

0,0319

86875

238500

40700

58,0

15,46

53,1

9,12

9,11

10,89

12,38

35,5

0,0283

91875

270000

32098

53,9

18,16

59,5

8,46

8,47

11,62

14,84

35,5

0,0248

95625

288000

24975

50,5

19,61

60,1

7,95

7,93

11,94

17,01

34,4

0,0212

99375

304500

18130

46,2

20,91

59,9

7,28

7,26

12,02

18,89

32,7

0,0177

101875

319500

12580

42,8

22,18

57,3

6,72

6,72

11,91

20,48

29,8

0,0142

104375

330000

8140

31,9

23,00

52,2

6,15

6,14

11,67

21,78

25,9

0,0106

106250

336000

4625

35,4

23,42

43,8

5,57

5,56

11,32

22,80

20,9

0,0071

107500

340500

2035

31,6

23,75

33,3

4,95

4,96

10,88

23,52

14,9

[%]

134

14

25 20

10 8

15

6

10

Head [m]

Power demand

12

4 5

2 0

0 20 %

30 %

40 %

50 %

60 % Flow

Akseliteho näytöltä Akseliteho by PumpSave Nostokorkeus by PumpSave

70 %

80 %

90 %

Nom. Flow

Laskettu akseliteho Nostokorkeus

Figure 89. Shaft power and head curves.

Shaft power curves in figure 89 show that values calculated by PumpSave differ drastically from measured and calculated power demand. This is caused by pump’s efficiency which is rather low when operating at full speed. Efficiency starts to increase rapidly when flow is throttled and reaches its maximum point at 70 % of the maximum flow. Efficiency input as an initial value for PumpSave does not get any better at any point of throttling but starts to decrease when flow is about 80 % of the maximum. Efficiency curves are illustrated in figure 90 and are the main reason to the difference between shaft powers. Figure 89 shows also that head curves differ from each other quite a bit. Difference between curves is larger when pump is operating at larger volume flows and decreases when operating point enters the area of smaller flow rates. In reality pump produces higher head than PumpSave calculates at almost whole operating range. In this case it reduces the difference of shaft powers.

135

0,7 0,6 Efficiency

0,5 0,4 0,3 0,2 0,1 0 20 % 30 % 40 % 50 % 60 % 70 % 80 % 90 % Nom. Flow Flow Pump efficiency

Pump efficiency by PumpSave

Figure 90. Efficiency curves.

Next is studied a situation which matches PumpSave’s calculation much better. Nominal operating point of the pump is equal to best efficiency point. In performed throttling control test best efficiency point was found at 70 % of the maximum flow. So this point is chosen to be the nominal operating point and at this point flow is throttled 10 % at a time. Results are shown in figures 91 and X where shaft power, head and efficiency curves are illustrated.

9

25

8 20

6 15

5 4

10

3 2

5

1 0

0 30 %

40 %

50 %

60 %

70 %

80 %

90 %

Nom. Flow

Flow Calculated shaft pow er

Shaft pow er by PumpSave

Head

Figure 91. Shaft power and head curves [m]

Head by PumpSave

Head [m]

Shaft power [kW]

7

136

In this case values for shaft power are quite well matching between calculations. Shaft power red from the panel of the frequency converter matched the calculated power demand perfectly so it was left a side from the figure 91. Head curves are also almost identical in this case as can be seen from figure 91.

0,7 0,6

Efficiency

0,5 0,4 0,3 0,2 0,1 0 30 %

40 %

50 %

60 %

70 %

80 %

90 %

Flow Pump efficiency

Nom. Flow

Pump efficiency by PumpSave

Figure 92. Efficiency curves.

Efficiency curves for this case are illustrated in figure 92. Now the curves are considerably closer to each other through the whole operating range than previously. Naturally this contributes to the good results gained for shaft power.

Conclusions

Results of laboratory testing are quite similar to results gained from comparing PumpSave’s calculation to characteristic curve based calculation. PumpSave’s calculation works correctly when the pump’s nominal operation point equals the best efficiency point. In variable speed control PumpSave does not take into account any efficiency changes during rotation speed adjustment. Based on first introduced test run, pump’s efficiency changes through the operation range quite a bit. When setting the nominal operating point equal as the best efficiency point the results for efficiencies match PumpSave’s calculation better. Program could be installed with some correcting factors for pump’s efficiency at different flow rates but this would require more testing and studying of different types and sizes of pumps in laboratory and with characteristic curves. These kind of correcting

137

factors PumpSave and FanSave uses are always some kind of a compromise among different sizes of applications. For throttling control the results from laboratory test runs are well matching when the nominal operation point where the throttling starts equals the best efficiency point. PumpSave’s calculation does not match the results at all if the throttling is begun at the descending area of the efficiency curve. When using PumpSave to observe the power demand and energy consumption of different applications the user should be aware how the pump’s efficiency behaves when moving on to some other operating point than the nominal point. It is also good to remember that the characteristic curves for each pump are based on an ideal situation and in real life those values are hard or impossible to reach.

12 SUMMARY For powerful fans it would be good to include FanSave with possibility to calculate change in density and in temperature of gas during blowing. Change in gas’s temperature affect the power demand of the fan when change is big enough. Adding this feature into the program would require tests with different sizes of fans. At different flow rates density changes different amount so if the changes to the program would be made with constant factors which FanSave uses a lot different types of fans should be tested to find the average factors valid for various fans. One way could also be to add the option to input as initial values the temperatures for incoming and outgoing flow. FanSave could then solve the average density based on these values. FanSave’s manual does not contain any information about the SFP. Also the calculation of SFP in FanSave is incorrect. This could easily be corrected to match calculation presented in equation 17. Also the manual of the program should be added information about the SFP number.

138

Previous studies for different sizes of pumps controlled by a variable speed drive prove that PumpSave’s calculations are almost perfectly accurate. Power demands and heads for each cases are almost identical for each pump case. For some cases pump’s efficiency changes several percentages during the flow adjustment and has affect on the power demand. Especially when operating point enters the lowest flow rates PumpSave assumes that in variable speed control pump’s efficiency remains as a constant. Motor’s efficiency differs from PumpSave’s calculation when it is calculated with equations presented in chapter X. through the operating range for each pump motor’s efficiency calculated by PumpSave is worse than efficiency calculated using pump’s characteristic curve. If PumpSave’s calculation would be changed to match calculation presented in equations 41, 42, 43 and 44 more information about the driving motor’s operation would be required. At least torque and rotational speed at each operation point should be calculated.

12.1 Observation of the profitability of investment FanSave and PumpSave examines the investment of the variable speed drive with present value which has been subtracted the investment cost. If the present value is positive the investment is profitable. Present value is calculated with equation n

NA = ∑ t =1

St

(1 + i )

t

+

JA

(1 + i )n

−H

where i

interest rate

n

utilization time [a]

t

time

St

annual net savings or profit [€]

JA

residue value [€]

H

investment cost [€]

(78)

139

Payback period is calculated directly as the quotient of the investment cost and the annual saving. This method is called interest free payback period. Particularly for the larger investments it would be appropriate to calculate the payback period with interest. This could be calculated by using equation 78 or 79. This feature could also be added to both of the programs. All the factors are already included. np

St

∑ (1 + i ) t =1

t

−H =0

(79)

where np

payback time with interest [a]

⎛1 H ⎞ − ln⎜⎜ − ⎟⎟ − ln(i ) ⎝ i St ⎠ n= ln (1 + i )

(80)

140

REFERENCES

Electric Ideas Clearinghouse. North Seattle Community College Two-Speed Motors Case Study. 1993. [Viitattu 20.11.2007] Saatavilla www-muodossa: http://www.p2pays.org/ref/36/35282.pdf Fan Types. Environmental Protection Agency. [Viitattu 9.10.2007] Saatavilla www-muodossa: http://www.epa.gov/eogapti1/module5/fans/types/types.htm Fläktwoods Oy. SFP-opas. [Viitattu 21.11.2007] Saatavilla www-muodossa: http://www.flaktwoods.fi/476d6be3-be6e-42e9-bd82-6152ff71a7aa Goulds Pumps. Pump Selection System. [Viitattu 5.2.2008] Saatavilla wwwmuodossa: http://www.gouldspumps.com/pss.html Hirvonen Sampsa. Taajuusmuuttajan vaihtaminen pakkaajarobottiin. Insinöörityö. Helsingin ammattikorkeakoulu Stadia. Sähkötekniikka. 2007. Inlet and outlet dampers for centrifugal fans, 1999, [Viitattu 1.10.2007] Saatavilla www-muodossa: http://greenheck.com/pdf/centrifugal/centrifugal_damper_cat.pdf Inlet vane controls & dampers of any type, [Viitattu 1.10.2007] Saatavilla wwwmuodossa: http://www.damper-designs.co.uk/ Invertteri. Lappeenrannan teknillinen yliopisto. Sähkötekniikan osasto. [Viitattu 23.9.2007] Saatavilla www-muodossa: http://www.ee.lut.fi/taajuusmuuttaja.html Jokinen Markku. Induktiomoottorikäytön paikkasäädön suorituskyky. Diplomityö. LTY. Sähkötekniikan koulutusohjelma. 2004 Karjalainen Veikko. Huoltokoulu osa 10: Diodit, zenerit ja triakit. [Viitattu 29.2.2008] Saatavilla www-muodossa: http://physics.uku.fi/studies/kurssit/ELE1/luennot/hk_diodit.pdf

141

Karttunen Erkki. 2004. RIL 124-2 Vesihuolto II. 1. edition. Helsinki: Suomen Rakennusinsinöörien Liitto RIL ry. 314 pages. ISBN 951-758-431-8 Luentomoniste 1 Virtaus- ja lämpövoimakoneet, kompressorit, puhaltimet, pumput. 2004. LTY Luukkanen Petteri. Pumpunvalitsimet integroidussa toimintaympäristössä. Diplomityö. LTY. Kemiantekniikan osasto. 2001. Montonen Arto. Taajuusmuuttajalla syötetty oikosulkumoottori valssilaitosten kelauskäytössä. Diplomityö. Lappeenrannan teknillinen korkeakoulu. Energiatekniikan laitos. 1985. Partanen Jarmo. Sähköenergiatekniikan perusteet. Opetusmoniste EN C-98. 1997 Puhaltimet teollisuudessa ja LVI-laitoksissa. Insinöörijärjestöjen koulutuskeskus. 40. painos. 1986. INSKO ry. ISBN 951-794-939-1 Pumpputekniikka – nesteiden pumppaus. Insinöörijärjestöjen koulutuskeskus. 35. painos. 1981. Insinööritieto Oy. ISBN 951-793-513-7 Ruuskanen Anne. Optimization of energy consumption in wastewater pumping. Diplomityö. LTY. Energia- ja ympäristötekniikan osasto. 2007. Sarkomaa Pertti. 1997. Pumppujen taloudellinen käyttö. Lappeenrannan teknillinen korkeakoulu. ISBN 951-764-110-9 LVI-talotekniikkateollisuus ry. SFP-opas. 2. painos. 2005. Viitattu [20.12.2007] Saatavilla www-muodossa: http://www.teknologiateollisuus.fi/files/9727_sfpopas2_200705.pdf

142

SFS ISO 5167-1:2003 Measurement of fluid flow by means of pressure differential devices inserted in cicular cross-section conduits running full. Part 1: General principles and requirements. SFS ISO 5167-2:2003 Measurement of fluid flow by means of pressure differential devices inserted in circular cross-section conduits running full. Part 2: Orifice plates. Seppänen Olli. 1988. Ilmastointitekniikka ja sisäilmasto. LVI-Kustannus Oy. ISBN 951-96098-0-6 Systemair. Teoria. [Viitattu 16.12.2007] Saatavilla www-muodossa: http://www.systemair.fi/data/attachments/teoria2.pdf Varttinen Sami. Taajuusmuuttajat voimalaitosten pumppauksissa, erityisesti syöttöveden pyörimisnopeussäädön vaikutus ruiskutusvesijärjestelmiin. Diplomityö. LTY. Energiatekniikan koulutusohjelma. 2004. Ventur Centrifugal fan GMT. [Viitattu 29.11.2007] Saatavilla www-muodossa: http://www.ventur.fi/content/att/06_GMT_eng.pdf Viholainen Juha. Developing of the Variable speed drive –control in simulated parallel pumping systems. Diplomityö. LTY Energia- ja ympäristötekniikan osasto. 2007. Volk Michael. 2005. Pump Characteristics and Applications. 2. edition. CRC Press, Taylor & Francis Group. Torque Limiters and Slip Clutches Information on GlobalSpec. GlobalSpec. [Viitattu 9.10.2007] Saatavilla www-muodossa: http://motioncontrols.globalspec.com/LearnMore/Motion_Controls/Clutches_Brake s/Slip_Clutches?SrchItem=1&frmqry=slip%20coupling

143

144

Appendix A: Mainpage of FanSave

145

Appendix B: Mainpage of PumpSave

146

Appendix C: Discharge coefficient

147

Appendix D: Expansibility factor

148

Appendix E: Centrifugal fans Appendix E1: HF R 560-13 D/R

149

Appendix E2: HF R 200-48 R

150

Appendix E3: HF R 710-13 D/R

151

Appendix E4: HF R 1000-13 D/R

152

Appendix E5: GTLF-3-025

153

Appendix E6: GTLF-3-040

154

Appendix E7: GTLF-3-056

155

Appendix E8: GTLF-3-071

156

Appendix E9: GTLB-3-025

157

Appendix E10: GTHB-3-040

158

Appendix E11: GTHB-3-056

159

Appendix E12: GTHB-3-071

160

Appendix E13: GTHB-3-100

161

Appendix E14: GTLB-3-140

162

Appendix F: Axial fans Appendix F1: HF A 630

163

Appendix F2: HF A 800

164

Appendix F3: HF A 1000

165

Appendix F4: Applied Energy XLCA1000

166

Appendix F5: Ebm-Papst 3G400

167

Appendix F6: Ebm-Papst 3G990

168

Appendix F7: Ebm-Papst 3G800

169

Appendix F8: Ebm-Papst 3G650

170

Appendix F9: Ebm-Papst 3G710

171

Appendix G: Pumps Appendix G1: Goulds CDS 5309-1 model 3298

172

Appendix G2: Goulds CDS IC307-1 model IC/ICB/ICP/ICV

173

Appendix G3: Goulds 3638

174

Appendix G4: Goulds 3628

175

Appendix G5: Goulds 5500

176

Appendix G6: Goulds 3635

177

Appendix G7: Goulds 3652

178

Appendix G8: Goulds CWX

179

Appendix G9: Goulds 5500

180

Appendix G10: Afton ILVS 1.5x2-7A

181

Appendix G11: Afton ILVS 1.5x2-9H

182

Appendix G12: Afton ILVS 1.5x3-11H

183

Appendix G13: Afton ILVS 2x3-13H

184

Appendix G14: Afton ILVS 3x4-10

185

Appendix G15: Afton ILVS 4x6-11H

186

Appendix G16: Afton ILVS 3x4-13L

187

Appendix G17: Afton ILVS 4x6-13H

188

Appendix G18: Serlachius

189

190