The pump affinity laws are the three simple relationships that tell you how a centrifugal pump's flow, head, and power change when you speed it up, slow it down, or trim its impeller. They are the reason a variable frequency drive can save so much energy compared with throttling a valve, and the reason a small cut in speed makes a large cut in power. This guide explains the three laws, why the power relationship is so dramatic, and how the speed signals behind them show up in SCADA.
Pump Affinity Laws in one line: The pump affinity laws are three proportionality rules for centrifugal pumps: flow changes in direct proportion to shaft speed, head changes with the square of speed, and power changes with the cube of speed. So halving the speed roughly halves the flow but drops the head to a quarter and the power to an eighth. The same relationships apply, approximately, when an impeller is trimmed to a smaller diameter, which is why speed control and impeller trim are the two standard ways to rematch a pump to a new duty.
The first affinity law says flow is proportional to speed. Turn the impeller twice as fast and it sweeps twice the volume per unit time, so the flow rate doubles. This is the most intuitive of the three because it follows directly from the impeller moving fluid in proportion to how quickly it spins.
The second law says head is proportional to the square of speed. Head comes from the velocity the impeller imparts to the fluid, and that velocity rises with speed, so the pressure energy - which depends on velocity squared - climbs much faster than flow. Double the speed and the head goes up roughly fourfold. This is why a modest speed increase can push a pump against a much higher discharge pressure than the flow change alone would suggest.
The third law says power is proportional to the cube of speed, and it is the one with the biggest practical consequences. Power is the product of flow and head, and since flow scales linearly while head scales as the square, the two combine so that power scales as the cube. Double the speed and power demand goes up roughly eightfold; cut the speed by twenty percent and power falls by nearly half. These same relationships hold, approximately, when an impeller is trimmed to a smaller diameter instead of slowing the shaft, though trim is a permanent change while speed can be varied continuously.
When a process needs less flow than a fixed-speed pump delivers, the traditional fix is to close a control valve and throttle the excess. This works, but it is wasteful: the pump still spins at full speed drawing near-full power, and the valve simply burns off the surplus head as heat and turbulence. The energy thrown away across the valve is money spent moving pressure that the process never uses.
A variable frequency drive avoids that waste by slowing the pump instead of choking its output. Because of the cube law, reducing the pump to the speed that produces exactly the flow the process wants cuts the power demand dramatically rather than proportionally. A pump running at seventy percent speed to deliver seventy percent flow draws only around a third of full power, whereas the same flow achieved by throttling would still draw close to full power. Over a year of continuous operation on a variable-demand service, that difference dominates the pump's operating cost.
There is a limit to the savings, though, because most real systems have static head - a lift or backpressure that does not fall with flow. Slowing the pump reduces the friction component of the demand but not the static component, so the cube-law benefit is largest on friction-dominated systems and more modest where static head is high. Impeller trim gives a similar rematching effect without a drive, but it is fixed once cut, so it suits duties that will not change rather than services with swinging demand.
On a modern pump skid the speed that drives the affinity laws is not a hidden mechanical setting - it is a live value in the control system. A variable frequency drive publishes its output frequency, or a derived speed percentage, as a tag, and that single number lets an operator reason about how the pump's flow, head, and power should be behaving through the affinity relationships. Reading speed alongside flow and discharge pressure closes the loop between the drive command and the hydraulic result.
A cloud SCADA such as Merobix can collect the drive speed, motor current or power, flow, and discharge pressure together and trend them on one screen. Because the affinity laws predict how power should track speed, an operator can spot when the relationship breaks down - if the drive slows but power does not fall as steeply as the cube law implies, the pump may be fighting more static head than expected, or a valve downstream may be partly closed and defeating the whole point of the speed reduction.
The speed tag also feeds energy accounting. Trending drive frequency against flow over weeks shows how much of the day the pump spends at reduced speed and, through the cube relationship, roughly how much energy the variable speed operation is saving versus a fixed-speed baseline. That turns the abstract affinity laws into a concrete efficiency story that operators and managers can watch, rather than a formula that lives only on a datasheet.
Flow is proportional to speed, head is proportional to the square of speed, and power is proportional to the cube of speed. So if you change a centrifugal pump's speed, flow changes by the same factor, head changes by that factor squared, and power changes by that factor cubed. The same relationships apply approximately when an impeller is trimmed to a smaller diameter.
Because power scales with the cube of speed, slowing a pump to deliver less flow cuts its power demand far more than proportionally. Throttling a valve instead leaves the pump at full speed and full power while wasting the surplus head as heat. On a friction-dominated system the cube-law savings from speed control can be very large, though a high static head reduces them.
Yes, approximately. Trimming an impeller to a smaller diameter rescales flow, head, and power much like a speed change, following the same proportional-square-cube pattern. The main difference is that a trim is a permanent mechanical change, whereas a variable frequency drive can vary speed continuously, so trim suits fixed duties and drives suit services with changing demand.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.