Pump System Calculations
Pump System Calculations
Pump System Calculations
0 March 2009
This workbook allows the user to change flows, the pressure setpoint, and the reset schedule. This will allow the user
to estimate the impact of changes such as
• Reduce Differential Pressure setpoint
• Reset Differential Pressure setpoint
• Reduce flows (e.g. by changing 3-way valves to 2-way valves)
• Reduce operating hours (with schedule changes or temperature lockout)
This workbook includes some checks of performance relative to the requirements of California "2005 BUILDING
ENERGY EFFICIENCY STANDARDS" (Title 24). These checks are provided as a service to users. However,
performance estimated by this workbook is not warranted as, and should not be construed as, any proof of
compliance or non-compliance.
General Comments
There are 3 input worksheets: "PumpInput," "SystemInputs," and "CalcInputs." Data should be entered in these
worksheets in that order. The "InputDataGuide" tab has helpful guidance on obtaining or estimating the data that
needs to be entered.
If you don't have a specific pump curve to enter, the program will scale a default pump curve based on the design
sizing for the project pump.
This workbook simulates a simple flow loop. Multiple take-offs are not simulated. Therefore, it is important that the
user simulate the critical zone or coil in their system to properly estimate the energy use and savings. The critical
zone is the zone that requires the the controlling device (valve or damper) to be the nearest to fully open in order to
satisfy the zone. If a non-critical zone is simulated, the savings could be over-estimated. In the real system, if a non-
critical zone controls the duct or pump pressure, then a more critical zone could end up "starved" for flow. This will
often result in a controls change that will cause a reduction in savings from what was estimated.
Instructions Page 1
Systems with take-offs upstream of the static pressure sensor will give conservative results using the simple inputs on
worksheet "System Inputs." The degree of conservatism is dependent upon the proportion of the total pump flow that
takes off before the sensor. If more accurate results are desired, the more detailed inputs can be used. The more
detailed inputs allow the user to enter different flows at different places in the loop, but also require more information
about the pressures in the loop. If the flows are the same throughout the loop, and if the appropriate inputs are used,
simulations using the simple and detailed inputs will produce the same energy use and savings estimates.
Reset of the controlling differential pressure is simulated by allowing the user to input different pressures at different
flows. It is up to the user to input pressures that would result from a particular reset schedule or reset control strategy.
This workbook does not directly simulate a reset based on valve position. However, the program does provide a
formula that can determine the lowest possible pressure that can allow the input flow to be achieved. In other words,
this formula, which can be input automatically by a macro button, estimates the pressure that will cause the controlling
valve to be fully open and the flow still achieved..
This calculation requires an interative solution. The standard Excel add-in, "Solver," must be installed. If
Solver is not installed, click Add-Ins on the Tools menu, and then select the Solver Add-in check box. Click OK, and
Excel will install the Solver. If you don't see Solver on the list of possible add-ins, you may need to run Excel Setup
again to make it available.
There is a button on the "VFDCalcInputs" worksheet to run the interative calculations. This macro updates all of the
calculation worksheets.
If you don't have a successful solution initially, it is probably because the system pressure drops are too great for the
input flows and the specified pump. The application provides prompts for this circumstance. Also see the notes for
worksheet "SystemInputs."
IMPORTANT: The Excel Solver usually does a good job of making sure that the pressure drops at the input flow
match the pressure rise across the pump, and making sure that the input pressure setpoint is met. However, as with
all analyses, the user should review the output to make sure the results make sense. Watch for unexpected
discontinuities in the graphs, and verify that the pressure matches the setpoint.
Also, note that the application assumes that the required flow will be achieved unless it is impossible to achieve it,
even with the speed at 100%. In some cases this could mean that the sensed pressure is higher than the setpoint.
Clearly this is not the way the systems would really operate: in the field the speed would be decreased to satisfy the
pressure setpoint, and the required flow wouldn't be achieved. The alternative would have been to allow the flow to
vary, but this would have slowed the calculation and made it less stable. If the results show a sensed pressure higher
than the setpoint, check the inputs and make sure they truly represent the system being modeled.
Advanced Users. The sheets have been protected to avoid erroneous input. Advanced users who wish to inspect
the calculation method in detail or make changes can access the formulas and code in the sheet as follows: Unhide
the sheet named "Version." You may wish to add comments there about changes you make to your version of the
tool. At the top of the "Version" sheet are convenience buttons that will unprotect or protect the sheets. It is
recommended that the tool be put in protected mode for use to avoid overwriting necessary formulas.
Inputs
Input cells are shaded yellow.
Instructions Page 2
Worksheet "PumpInput"
This worksheet is used to input the Pump data.
Enter the design selection point for the pump, usually available from the pump nameplate or drawings. If the system
has 2 identical pumps controlled in parallel, the flow for one pump should be entered, half of total system flow. Note
that a 2-pump system may be designed for only one pump to operate, with the second pump as a backup and normal
entry would be used. Note also that building operators will often have both pumps running when only one is needed.
If both pumps are found on, the system design and control sequence should be investigated. If only one pump is
needed, sequences should be revised or building operators educated.
Pump curve input: Ten points must be input for both the head curve and the efficiency curve. The Pump speed is also
input.
The data must be entered in order, from lowest flow to highest flow. The lowest flow should be near zero, and the
highest flow should give a Pump head near zero. This will often need to be extrapolated from the manufacturer's
curve for entry here, since Pump curves typically don't cover such relatively high flows.
Though this workbook has auto-scaling of a default Pump curve, it is strongly recommended to input the actual Pump
curve or the curve for another similarly-sized Pump in order to gain the best results. The range of flow and head must
include the design pump flow and head. When the pump curve does not match the actual pump, expect the solver
iteration to require multiple attempts.
Worksheet "SystemInputs"
The first set of inputs are for the differential pressure setpoint, motor and VFD sizes (to get part-load efficiencies), a
choice of methods to get VFD efficiencies (see chart at right for comparision of one case), and a choice of using
Temperature or Flow as the category axis on one of the output charts on the "SavingsSummary" worksheet. The two
required inputs are motor size and VFD size (usually the same as motor size).
The next sets of inputs are used to describe the pressure drop characteristics of the system. The "Simple Inputs" are
easiest to use, and provide the same energy use and savings estimates as the "Detailed Inputs."
- Input tested pump pressure difference from points 1 to 2 on schematic and pressure differential measurement from
points 3 to 6, as well as measured or estimated flow during test.
- "Detailed Inputs" have the advantage of providing a better estimate of where the pressure drops occur in the
system, and hence more accurate reporting of pressures at particular places in the system. However, "Simple Inputs"
should suffice in the vast majority of cases.
If using the "Detailed Inputs," each set of 3 inputs—2 pressures and 1 flow—must have a consistent set of inputs: The
input pressures must correspond to the pressures at the input flow.
Different sets of inputs can be at different flow conditions, but the pressures must be the ones that result at the input
flow.
(A later version may permit use of a different flow for the critical zone.)
These inputs can come from trend data, measurements, drawings, or calculations as available. But the relationship of
the pressure drops in the system is key to estimating the savings when there is no differential pressure reset, and will
affect the savings of differential pressure reset whenever the reset is not perfect. Therefore, the inputs for flows and
pressures are important.
Instructions Page 3
Ideally the inputs would come from trend data or measurements. However, these data may be hard to obtain. If
measured data are not available, the system data for "∆p across pump" and "estimated flow" can be based on
the pump design data, adjusted for the assumption that that pump has some design margin. If there is no
design margin, this may be reflected in the calculations, and the baseline calculation could fail because it
might not be possible to achieve the design flow.
Worksheet "VFDCalcInputs"
In these worksheets the required flows to handle the load at various operating conditions must be input. Flows at
different loads can be derived from trend data or energy simulation programs or methods. If the system has 2
identical pumps controlled in parallel, the flows for one pump should be entered, half of total system flow.
Up to 20 different flow conditions can be input. These will typically be for different ambient conditions, and a column
of bin temperatures are provided for convenience. However, the flows could vary based on other criteria as well. It is
up to the user to create a correlation of flow to the appropriate independent variables. The bin temperature column
can be edited as needed. The bin temperatures aren't used by the program; they are only there for
convenience in listing the hours at a particular flow. The user could use the bin temperatures in a formula to
calculate the flows, if the flows and load are dependent on ambient temperature.
If the base case is fixed speed, then input a high (impossible to reach) pressure setpoint for baseline. If a VSD is
already in place, enter the baseline differential pressure control setpoint.
The "Change Flows and operating hours" columns enable the user to simulate system changes that would change the
flows required to satisfy the load. For example, the flow could change if the minimum flows for one or more coils
changed. The flows could also change if the load was reduced as part of the project, such as by a lighting retrofit, or
replacing some 3-way valves with 2-way valves.
Operating hours can be reduced through schedule changes, reducing building loads, or putting a high- or low-
temperature lockout in to turn off the pump when not needed.
The user can enter the specific differential pressure setpoints associated with a reset schedule.
(Range K8:K27) These cells can include a formula that uses the pressure drop characteristics of the piping and
equipment to simulate a "perfect" reset schedule. That formula is as follows:
The formula also has logic to maintain a minimum pressure differential as input.
This formula will be placed in Range K8:K27 by clicking the macro button to "Use Optimum Reset Pressures"
The "perfect" reset uses a formula to give the theoretical optimum setpoint for each flow. Based on the input
pressures, the piping characteristic pressure drops are estimated, and a "perfect" reset pressure is the minimum
pressure needed to push the input flow through the simulated system. Because this application only uses a single-
zone model, and pressure drops may not always be, as simulated, a function of the square of the flow, the values
resulting from that formula will give optimistic results. Since the "perfect" resets are not likely to be achieved in reality,
they should just be used to see the limits of what might be gained with setpoint reset. In addition to zone differences,
the theoretical optimum reset shown assumes the abilitiy to use a valve-position based reset strategy that may be
difficult as discussed in the Seidl article on the links page.
Hint: If the calculation will not complete, first run a case where you pick a design point from the representative pump
curve provided to prime the cells with normal values for the solver, then rerun with the actual pump design conditions
that require an extrapolated pump curve.
Calculation Worksheets
Instructions Page 4
Calculation worksheets: "CalcBaseline," "CalcReduceFlows," "CalcWithVFD," and "CalcWithReset."
These worksheets are where the iteractive calculations take place. The calculations solve for the required pressures
in the flow loop. No user inputs are required. If you wish to stop a calculation underway, use the ctrl key and "break"
button several times.
Note that the current version does not include a full documentation commentary of all formulas, although they are
fairly standard engineering calculations that should be understandable to an experienced HVAC engineer.
Calculation Methodology
The basic approach is to use the excel solver application that applies a Newton-Raphson convergence methodology
on the calculation sheets. Each calculation sheet is essentially identical, using different inputs for the relevant
situation. Relevant data are taken from the input sheets and the solver iterates values for the pump speed and critical
throttling valve until a solution is reached for each flow rate. Note that the solver writes unique case values into
the sheet for pump speed and critical valve pressure drop, so until the calculation button is pressed, the
values will not be correct for a new situation that has been entered. If detailed inputs are provided on the
SystemInputs tab they are used, otherwise intermediate system pressures are estimated based on simplified inputs.
Parameters for each pressure position shown on the schematic are shown on the SystemInputs tab below the
schematic.
Once pump speed and pressure is calculated for each flow bin, the input efficiencies as projected to the selected
pump curve are used along with standard motor efficiency factors and the selected VSD power determination method
to calculate pump kW and pump kWh at each flow bin.
Output Worksheet
Worksheet "SavingsSummary"
This worksheet should be self-explanatory. If the system uses parallel pumping, with both pumps operating at all
times, savings will double the value shown.
Other Worksheets
Worksheet "PumpCurveFit"
This worksheet estimates two polynomial fits to the input Pump data, a 4th-order fit for the pressure and a 3rd-order fit
for the efficiency.
These curve fits are used to estimate Pump performance in the Calc worksheets.
The worksheet also has a graph showing the Pump curves and some cells generalizing the curve so that it can be
used for different sized Pumps.
Worksheet "VFDeffy"
The user doesn't need to do anything with this worksheet.
This worksheet shows the sources of the data for motor and VFD efficiencies.
Problems?
Please email the California Commissioning Collaborative at info@cacx.org, if you encounter problems with this tool.
Instructions Page 5
Links to Related Resources
Pumps and Your Processes: They Must Work Together
Using System Curves to Enhance Pump Performance
Pumps and Your Processes: When They Won't Work Together
Continuous Commissioning Guidebook (FEMP)
CCSM Measures for Water/Steam Distribution Systems
Workshop Report on DDC to the Zone Level
David Sellers' Blog
Right-Sizing Pumps, David Sellers, HPAC Magazine March 2005
2005 CA Building Energy Efficiency Standards website
2005 CA Building Energy Efficiency Standards Manual
2005 CA Nonresidential Compliance Manual
Using demand-based reset strategies, Reinhard Seidl
Data Inputs are color coded as follows: Required Data (enter site-specific data in yellow cells)
Note: Data is NOT input on this sheet! Optional Data (leave defaults if acceptable)
The following data is needed for this calculation. Detailed Input Data or rarely overriden default calculations
Please see the reference guide downloaded with the tool for recommended testing procecures to acquire some of this data.
Input Data Units Input on Tab Preferred Source
pump conditions at peak efficiency
pump Speed rpm pumpInput Pump curve for investigated pump at
hydronic Flow GPM pumpInput design speed. Where available, put in
Head ft. Head pumpInput actual pump efficiency.
Max Efficiency % pumpInput
Above for Design pumpInput Pump nameplate tag, equipment
submittals
pump curve data (default pump pumpInput pump curve for main supply pump at
curve will scaled to specific pump design speed.
inputs, so not high importance.)
Use Detailed Inputs Check SystemInputs Check only if have all noted test data or good estimates
differential Pressure psi SystemInputs Field test of pump flow and pressures
at noted points with all coil valves full open
Main pump GPM SystemInputs
hydronic flow
Initial Flow Factor SystemInputs Recommended to leave this at 1.05
Operating hours at hours/yr VFDcalcInputs Flow trends correlated with outside air
each load condion temperature or load conditions.
Plans or submittals
Plans or submittals or estimate
same as motor size
Bernier method has typically lower
efficiencies at low flow and higher
energy use than DOE-OIT method
Used in chart titles only
Estimated design flow and head If the differential pressure control does
conditions (calculate differential not exist or is located too close to the
across points [3-6] using pipe pump, it may be beneficial to test
presure drop calculations). If differential pressure in hydraulically
measurements cannot be extreme locations to justify installation
obtained for full flow differential or relocation of differential pressure
pressure (point [3-6]), care must sensing.
be taken in estimating this value.
Simulation of building with hydronic Required for Base and Reduced flow
flow trends by load condition, or or hours Case
typical estimate based on similar
building.
Manufacturer's recommendation of Note: While pressure independent
minimum valve inlet pressure for valves typically require 5 psi to
proper operation. provide accurate flow information, if
valve flow information is not
necessary a lower minimum setpoint
can be used and resets can be
controlled from valve position.
matic for reference:
Project Pump Info Speed, rpm Flow, gpm head, ft effy ∆p, psi
Maximum efficiency point 1750 135 120 56.1% 52.2
Project Pump Design Point 1750 120 125
Design point is optional entry,
for reference only
Note: Hover over comment marked (red triangle) cells for more detail on inputs.
Pump Curve: Generic (from B&G Series 80 2x2x7 @ 3500 rpm)
Impeller dia., in. Speed, rpm Flow, gpm head, ft effy ∆p, psi Water HP
7 3500 0 229.0 0.1% 99.6 5.8
44 228.0 30.0% 99.1 8.5
75 225.0 45.0% 97.8 9.5
106 221.0 54.0% 96.1 11.0
135 212.0 58.0% 92.2 12.5
170 198.0 60.0% 86.1 14.2
190 188.0 58.6% 81.7 15.5
210 170.0 55.5% 73.9 16.3
260 130.0 40.0% 56.5 21.4
330 0.3 0.1% 0.1 25.1
Pressure Rise, feet
Pressure Rise at
120 80% Speed
80 Design Point
60
40
20
0
0 50 100 150 Flow,200
gpm 250 300 350 400
Cases analyzed:
Baseline - Constant Flow
Details for case 0
Power [kW]
10
Improved: Valve controls flow; VFD controls ∆p
#VALUE! Remove pipe obstructions energy use
#VALUE! Additional savings to the reduced flow & hours scenario 8 With reset of ∆p setpoint
Efficiency
Baseline System Curve 80%
Reduced Hours
120 Improved System Curve
Improved: Valve controls Improved with Reset 70%
8 flow; VFD controls ∆p
Efficiency at 100% Speed
100
With reset of ∆p setpoint Efficiency at 80% Speed 60%
6 80 50%
40%
60
4 30%
40
20%
2 20
10%
0 0%
0 0 50 100 150 200 250 300
Ambient Temperature [ºF] Flow [gpm]
97.5 122 35 35 1750 47% 90% 90% 54 125 160 145 102 0.0 7.8 94
92.5 110 84 35 1750 42% 81% 93% 56 129 164 151 116 0.0 33.3 83
87.5 99 133 35 1750 38% 73% 94% 57 131 166 156 127 0.0 53.7 74
82.5 89 343 35 1747 34% 66% 95% 57 132 167 159 136 0.0 69.5 66
77.5 80 539 35 1719 31% 59% 96% 56 129 164 157 138 0.0 78.2 60
72.5 72 686 35 1697 28% 53% 97% 55 126 161 156 141 0.0 85.3 55
67.5 65 707 35 1679 26% 48% 97% 54 124 159 155 143 0.0 91.0 52
62.5 58 679 35 1664 23% 43% 98% 53 122 157 154 144 0.0 95.6 48
57.5 53 392 35 1652 21% 39% 98% 53 121 156 153 145 0.0 99.3 46
52.5 47 217 35 1643 19% 35% 98% 52 120 155 153 146 0.0 102.4 44
47.5 43 35 35 1635 17% 32% 98% 52 119 154 152 147 0.0 104.8 42
42.5 38 7 35 1628 16% 28% 99% 51 118 153 152 148 0.0 106.8 41
37.5 34 1 35 1623 14% 26% 99% 51 118 153 152 148 0.0 108.5 40
32.5 31 0 35 1618 13% 23% 99% 51 117 152 151 149 0.0 109.8 39
27.5 28 0 35 1614 12% 21% 99% 51 117 152 151 149 0.0 110.8 38
22.5 25 0 35 1611 10% 19% 99% 51 117 152 151 149 0.0 111.7 37
17.5 23 0 35 1609 9% 17% 99% 51 116 151 151 149 0.0 112.4 37
12.5 20 0 35 1606 8% 15% 99% 51 116 151 151 150 0.0 112.9 37
7.5 18 0 35 1604 8% 14% 99% 50 116 151 151 150 0.0 113.4 36
effective
net
exponent Regression: Pump power vs. flow Note: Regressions may be inaccurate if appropriate flows are not input for all 20 load conditions
87.5 84 133 35 1731 32% 62% 96% 57 130 165 158 137 0.0 74.4 63
82.5 76 343 35 1707 30% 56% 96% 55 127 162 156 140 0.0 82.2 58
77.5 68 539 35 1687 27% 50% 97% 54 125 160 155 142 0.0 88.5 53
72.5 61 686 35 1671 24% 45% 97% 54 123 158 154 143 0.0 93.6 50
67.5 55 707 35 1658 22% 41% 98% 53 122 157 154 145 0.0 97.7 47
62.5 50 679 35 1647 20% 37% 98% 52 120 155 153 146 0.0 101.1 45
57.5 45 392 35 1638 18% 33% 98% 52 120 155 152 147 0.0 103.8 43
52.5 40 217 35 1631 16% 30% 98% 52 119 154 152 147 0.0 106.0 41
47.5 36 35 35 1625 15% 27% 99% 51 118 153 152 148 0.0 107.8 40
42.5 33 7 35 1620 13% 24% 99% 51 118 153 151 148 0.0 109.2 39
37.5 29 1 35 1616 12% 22% 99% 51 117 152 151 149 0.0 110.4 38
32.5 26 0 35 1613 11% 20% 99% 51 117 152 151 149 0.0 111.3 38
27.5 24 0 35 1610 10% 18% 99% 51 117 152 151 149 0.0 112.1 37
22.5 21 0 35 1607 9% 16% 99% 51 116 151 151 149 0.0 112.7 37
17.5 19 0 35 1605 8% 14% 99% 50 116 151 151 150 0.0 113.2 36
12.5 17 0 35 1603 7% 13% 100% 50 116 151 151 150 0.0 113.6 36
7.5 16 0 35 1602 6% 12% 100% 50 116 151 151 150 0.0 113.9 36
Regression: Pump power vs. flow Note: Regressions may be inaccurate if appropriate flows are not input for all 20 load conditions
97.5 104 35 35 1496 46% 77% 90% 40 92 127 115 84 0.0 6.9 77
92.5 93 84 35 1448 43% 69% 92% 38 87 122 113 88 0.0 18.7 69
87.5 84 133 35 1408 40% 62% 93% 37 84 119 112 91 0.0 28.3 63
82.5 76 343 35 1376 37% 56% 95% 35 81 116 110 94 0.0 36.1 58
77.5 68 539 35 1350 34% 50% 95% 34 79 114 109 96 0.0 42.4 53
72.5 61 686 35 1330 31% 45% 96% 34 77 112 108 97 0.0 47.5 50
67.5 55 707 35 1313 28% 41% 97% 33 76 111 107 99 0.0 51.6 47
62.5 50 679 35 1299 25% 37% 97% 32 74 109 107 100 0.0 54.9 45
57.5 45 392 35 1288 23% 33% 98% 32 73 108 106 101 0.0 57.7 43
∆p
52.5 40 217 35 1279 21% 30% 98% 32 73 108 106 101 0.0 59.8 41
47.5 36 35 35 1272 19% 27% 98% 31 72 107 106 102 0.0 61.6 40
42.5 33 7 35 1266 17% 24% 98% 31 71 106 105 102 0.0 63.1 39
37.5 29 1 35 1261 16% 22% 99% 31 71 106 105 103 0.0 64.2 38
32.5 26 0 35 1257 14% 20% 99% 31 71 106 105 103 0.0 65.2 38
27.5 24 0 35 1253 13% 18% 99% 31 70 105 105 103 0.0 65.9 37
22.5 21 0 35 1250 11% 16% 99% 31 70 105 105 103 0.0 66.6 37
17.5 19 0 35 1248 10% 14% 99% 30 70 105 105 104 0.0 67.1 36
12.5 17 0 35 1246 9% 13% 99% 30 70 105 105 104 0.0 67.5 36
7.5 16 0 35 1244 8% 12% 99% 30 70 105 104 104 0.0 67.8 36
Regression: Pump power vs. flow Note: Regressions may be inaccurate if appropriate flows are not input for all 20 load conditions
Regression: Pump power vs. flow Note: Regressions may be inaccurate if appropriate flows are not input for all 20 load conditions