Grundfos varying head control does not need a signal representing valve-opening percentage. It changes the pressure target according to the pump’s inferred operating point. At high demand, the target approaches the commissioned full-demand head; as demand falls, the target moves toward the reduced-head endpoint. The key commissioning task is to establish endpoints that still satisfy the hydraulically remote consumer.
Control-mode prerequisites
- Confirm that the distribution system is closed-loop water circulation with two-way control valves. As valves close, flow falls and system resistance rises.
- Confirm where differential pressure is measured. A pump-mounted measurement represents pressure across or near the pump; a remote sensor represents pressure at its installed location. The required setpoint depends strongly on that location.
- Confirm whether the selected mode is constant head, varying head, or an adaptive mode. These modes must not be treated as interchangeable.
- Obtain the installed pump’s mode description. The product model is required to determine whether it estimates flow from speed and power, uses measured flow, or applies another internal calculation.
Do not move on until the displayed mode and pressure-feedback location match the intended control strategy. A controller configured for constant head will retain a horizontal pressure target even if the system was commissioned for a sloped target.
Full-demand reference check
Open the consumer valves to represent the maximum required demand, then balance the branches while adjusting pump speed. The full-demand endpoint is valid only when every required consumer receives its design flow and the remote branch has adequate differential pressure.
| Reading | Meaning | Next check |
|---|---|---|
| All consumers satisfied at the selected head | The full-demand pressure endpoint is usable | Check the reduced-demand endpoint |
| Remote consumer is starved | The head is too low, balancing is incomplete, or another hydraulic restriction exists | Correct flow and balancing before setting the endpoint |
| Large excess pressure at consumers | The endpoint may be higher than the network requires | Reduce speed carefully and recheck the remote consumer |
In the stated example, the resulting full-demand head is 10 m. That value becomes the upper endpoint. It is not automatically proven by valve position; it is established by hydraulic performance at the required operating condition.
Reduced-demand endpoint check
Varying head control lowers the pressure target as flow decreases. In the stated arrangement, the lower endpoint is one-half of the full-demand value, so a 10 m upper endpoint corresponds to 5 m at the low-flow end.
This reduction compensates for friction loss. Pipe and fitting losses fall sharply as flow falls, while the static pressure requirement of a closed circulation loop does not increase merely because the valves close. Holding 10 m at every flow can therefore impose unnecessary differential pressure across nearly closed valves, increasing valve noise, control sensitivity, and pumping energy.
| Low-flow observation | Interpretation | Action |
|---|---|---|
All active consumers remain controllable near 5 m
|
The reduced endpoint is hydraulically feasible | Test intermediate demand |
| A remote consumer loses authority or flow | The reduced endpoint is too low for the active path | Raise the lower endpoint or relocate the pressure sensor |
| Valves remain nearly closed with high differential pressure | The reduced endpoint may still be excessive | Lower it in controlled increments and retest |
Intermediate operating-point decision
The inclined line is a schedule of pressure target versus operating demand. Using normalized demand q, where zero represents the low-flow endpoint and one represents the full-demand endpoint, a linear schedule can be expressed as:
Hsp = Hlow + (Hfull - Hlow) × q
For the stated 5 m and 10 m endpoints, the target lies between those values whenever inferred demand lies between zero and full demand. The calculation does not mean that a valve position of 50% produces 50% system flow. Valve characteristics, branch pressure, simultaneous demand, and balancing all break that direct relationship.
The controller instead derives an operating point from the information available inside the pump system. Pump speed and developed head can locate an approximate point against stored pump characteristics; some implementations can also use motor power or a flow measurement. For a fixed impeller, the affinity relationships explain why speed is useful: flow varies approximately with speed and head approximately with the square of speed when comparing corresponding points. Speed alone, however, is not a universal measurement of building demand.
If the displayed pressure target moves between the endpoints as valves close, the varying-head schedule is active. If the target remains fixed, first check the selected control mode. If the target moves but remote pressure becomes inadequate, the endpoint selection or sensor location—not valve percentage—is the resolving branch.
Control-response and recurring pitfalls
A valve closure initially reduces flow and changes the differential pressure. The speed controller then reduces pump speed until measured pressure matches the newly calculated target. This creates two linked actions: the operating-point estimate changes the target, and the pressure loop changes speed to reach that target.
| Symptom | Likely cause | Diagnostic |
|---|---|---|
| Pressure stays at the upper endpoint | Constant-head mode selected, target schedule inactive, or estimated flow remains high | Record displayed mode, target, speed, and flow indication while closing valves |
| Pressure falls but the remote branch starves | Lower endpoint or slope is too low, or pump-mounted sensing does not represent the remote branch | Measure differential pressure at the remote consumer |
| Pressure or speed hunts | Valve action and pressure-loop response interact | Trend pressure target, actual pressure, speed, and valve command together |
| Expected target does not match valve percentage | Valve position is being mistaken for total demand | Compare actual or estimated flow with the target instead |
Do not tune the pressure loop before correcting an invalid endpoint. Loop tuning cannot compensate for a target that fails to provide the required remote differential pressure.
Commissioning and verification procedure
- Open the required consumers, balance the circuit, and reduce or increase pump speed until every design flow is met. Record the resulting head; in the example, it is
10 m. - Select varying head control and enter or confirm the upper endpoint. Confirm that the displayed target reaches the upper value at the full-demand operating point.
- Confirm the lower endpoint defined by the selected mode. For the stated half-head relationship, verify
5 mwhen the upper value is10 m. - Close valves in stages. At each stage, record actual differential pressure, displayed target, pump speed, flow or estimated flow, and remote-consumer performance. Do not move on until actual pressure settles at the displayed target.
- At low demand, verify that the target approaches the lower endpoint without starving the remote active consumer. If starvation occurs, raise the lower endpoint or revise the sensing location.
- Reopen the valves. Confirm that the target and speed rise smoothly, the head returns toward
10 m, and all required consumer flows recover.
FAQ
Does varying head control measure valve-opening percentage?
No. It normally acts on a measured or inferred pump operating point, so 50% valve travel must not be interpreted as 50% total demand.
Can I use half of the full-demand head as the minimum?
Use 5 m for a 10 m upper endpoint only when the selected control mode defines that relationship and field testing shows that the remote active consumer remains satisfied.
Does pump speed alone prove the system flow?
No. Speed becomes useful when combined with pressure and the pump characteristics; the installed model’s control description identifies any additional use of power or measured flow.
Can I verify varying head control without knowing its internal algorithm?
Yes. Trend target pressure, actual pressure, speed, and flow indication while closing and reopening valves; the final verification is smooth target movement between 5 m and 10 m with the remote consumer satisfied at every tested operating point.