Pump Head and Power
A pump buys head with power, and efficiency sets the price.
Before this Lesson
- The Energy Equation · Fluid Mechanics
- Power and Efficiency · Dynamics
A town's pump station draws 1,000 gallons a minute from a river and sends it up to a storage tank on a hill, where the water surface stands 150 ft above the river's. Picture doing that by hand. A 5-gallon bucket of water weighs about 42 lb, so 1,000 gpm is more than three full buckets carried up a 15-story building every second, all day and all night.
A pump does it with an electric motor, and the town pays the electric bill. By the end of this Lesson, you'll know how much power that motor draws and what a day of its work costs.
Weight per second, times height
Lifting a weight through a height takes work . A pump does that without stopping: every second, a fresh batch of water passes through it, and the pump raises it through the head . The weight passing each second is the specific weight times the flow, . Work per second is power, as in Power and Efficiency, so the power that ends up in the water is
The Handbook's Centrifugal Pump Characteristics section calls this the fluid power. Its head is the same as in The Energy Equation, and it writes power as , which keeps free for pressure.
From the Handbook's Units and Conversion Factors table:
- 1 ft³ = 7.48 gal, so 1 cfs = 7.48(60) = 448.8 gpm. Divide gpm by 448.8 to get cfs.
- 1 MGD = 1.547 cfs. The table prints it the other way, as cfs times 0.646317 to get MGD.
- 1 hp = 550 ft-lb/s = 745.7 W, or about 0.746 kW.
- 1 psi = 2.31 ft of water, which is 144/62.4.
In U.S. units, in cfs times 62.4 lb/ft³ times in feet gives ft-lb/s, and 550 of those make a horsepower. The Hook's 1,000 gpm is 2.228 cfs, so lifting it 150 ft takes
SI is kinder. With , in m³/s, and in meters, the product comes out in kN·m/s, which is kilowatts, with no conversion at all.
No pump hands all of its shaft's power to the water. Some churns into eddies in the casing, some drives water that leaks back around the impeller, and some goes into bearing friction, and it all ends up as heat. So the shaft must supply more than the water receives, and the ratio is the pump efficiency :
This is the Handbook's pump (brake) power, the power at the pump's shaft. At 75% efficiency, the Hook's pump needs there. Dividing by a number below 1 always makes it bigger, and that's your check: the shaft can't deliver less than the water gets.
Total dynamic head
So far the head was simply the lift. The energy equation says what a pump really has to supply. Put point 1 on the surface of the water the pump draws from, a sump or wet well, and point 2 on the surface of the tank it fills. Their pressure and velocity heads drop out, as in The Energy Equation:
The first term is the static head, the lift from one water surface to the other, there whether or not any water moves. The second is every loss the moving water pays: pipe friction, which a problem states or has you find by Hazen-Williams or Darcy-Weisbach, plus a minor loss at each entrance, bend, valve, and exit, with the fitting's loss coefficient. Those losses exist only while water flows, which is why the sum is called the total dynamic head (TDH). It's the of the power equation. The Handbook defines neither term, and "total dynamic head" appears there only as an axis label on its sample pump curve.
Nothing else in a pump station's data enters the static head. Slide the pump up or down its suction line, and the water still starts at the sump's surface and ends at the tank's. (The pump's height does decide whether its suction side can boil into vapor, a question this Lesson leaves out.) The floors of the sump and tank don't count, and neither does where the pipe enters the tank: water entering below the surface still has to push against all the water above it. Only a pipe that ends above the water and spills freely has to lift it to the outlet instead.
On a pump that's already running, there's a shortcut: read the head off its gauges. Write the energy equation from the suction flange, s, to the discharge flange, d, with nothing but the pump between them:
With the gauges at one height on pipes of one size, as problems usually arrange, the last two terms vanish, and the pump head is the jump in pressure head across the pump, at 2.31 ft of water per psi.
One sign needs care. A pump that sits above its sump has to pull water up its suction pipe, so the suction gauge reads below atmospheric, a vacuum. A vacuum is a negative gauge pressure, and subtracting a negative adds. A pump that takes water at a 4 psi vacuum and delivers it at 50 psi raises its pressure by 54 psi, which is of head.
From the wire to the water
The pump's shaft is turned by an electric motor, and the motor has losses of its own: its windings and bearings warm up. So the power it draws from the wire is larger again, by its efficiency :
This is what the Handbook calls the purchased power, and it's what the utility pays for. Each machine divides once, and as in Power and Efficiency, efficiencies in a chain multiply. The Handbook's Pump Power Equation packs the whole chain into one line, with a total efficiency :
Use it with for the motor input, or with the pump efficiency alone for the brake power.
The electric bill charges for energy, power times time, in kilowatt-hours: the motor input in kW, times the hours it runs, times the rate per kWh. Now finish the Hook. Say its pipeline loses 20 ft at 1,000 gpm, so the TDH is 170 ft and the water power is 43.0 hp. With a 75% pump and a 90% motor:
A day of that is , and at $0.12 per kWh, the town pays about $137 a day to lift its water.
Working backward
The same equation runs backward. A pump test measures the flow with a meter, the head with gauges, and the power with an electric meter or a shaft dynamometer. The efficiency is then what the water got, divided by what went in:
Measure the electric input, and this ratio is ; measure the shaft power, and it's alone. If the efficiency is known instead, solve the power equation for the flow a given power can move, or the head it can add:
Here multiplies, and that's no contradiction. The power going in, times the efficiency, is the share that reaches the water, and only that share lifts it.
A day's water in fewer hours
Many problems state a job rather than a flow: a town uses 2 million gallons a day, and its pumps must keep up. Pumps rarely run around the clock. A utility may run them only at night, when power is cheaper, or only during one shift. The day's volume is fixed, so fewer hours means a faster rate:
Pump 2 MG over 24 hours, and the rate is . Squeeze it into 8 hours, and the rate is 4,167 gpm, three times as fast, so at the same head it takes three times the power. Divide by the hours the pump actually runs, not the 24 in a day, and the power follows as before.
Static head stays put, friction doesn't
Pumping faster changes the head, too. Look again at the two parts of the TDH. The static head is a difference in elevation, so it doesn't care how fast the water moves: at a trickle or a flood, the lift is the same. The losses are different. Each grows with the velocity head, and the velocity grows with the flow. Darcy-Weisbach, , goes as for a steady , and Hazen-Williams as , close enough that problems usually say "roughly the square of the flow." Double the flow, and the friction head about quadruples, while the static head doesn't move.
Plotted against flow, the TDH is a flat floor of static head with the losses rising above it in a parabola. So the TDH isn't proportional to the flow, and the power, , climbs faster than the flow does, because the head climbs with it. To find the head at a new flow, scale only the losses:
Pump Curves and Operating Points calls this curve the system curve and finds where a pump will run on it.