Art 2-2002

Version 5 -9/09/02

1/6

Some applications in HVAC of self-acting differential pressure controllers

To enable stable and accurate control, differential pressures across modulating control valves should not
vary too much. This can be obtained with a self-acting differential pressure control valve STAP in a
variable flow distribution. This solution has some specific advantages:
1- Enable stable and accurate modulating control,
2- Minimise noise from control valves,
3- Avoid interactivity between circuits, which are maintained independent.

1- How it works

Fig 1: STAP stabilises the secondary differential pressure DpL.

STAP is a self-acting proportional control valve that stabilises the supply differential pressure of a circuit
or a branch with several circuits.
The design of STAP is based on a spring-membrane combination. The spring pulls the balanced plug
(2) to open the valve. Differential pressure AB is applied on the membrane (3), creating a force against
the spring. Pressure A is communicated to the STAP by means of a capillary connected to the drain (6) of
the measuring valve STAD/M (STAD or STAM). Pressure B is communicated internally to the other side
of the membrane.
The measuring valve may be cancelled and replaced by just a test point on the pipe. This is not
recommended if water flow is not measurable using another feature.
When the force created by the differential pressure AB on the membrane is higher than the force of the
spring, the valve starts to shut proportionally until it finds a new equilibrium position. This creates a
supplementary pressure drop in the STAP that limits the increase of differential pressure DpL.
The design force of the spring is modified with an Allen key introduced through the centre of the
handwheel (7). This allows the adjustment of differential pressure DpL to the required value. The
handwheel (7) itself is used to shut the STAP to isolate the circuit when necessary.
Water flow can be measured easily with the STAD/M and a balancing instrument CBI. Secondary
DpL is measurable between (5) and (4b), (when 4b is equipped with an optional test point) or between (5)
and (8).
As STAP is a proportional controller, differential pressure DpL is not maintained absolutely constant, but

Art 2-2002

Version 5 -9/09/02

2/6

Fig 2: The secondary differential pressure depends on the degree of opening of the STAP.

The Kv of the STAP varies from 0 to a maximum value Kvs. However the working range is situated
between a Kvmin to a Kvm value for which the secondary differential pressure takes stable values at
20% around the set point (SP). To achieve stable function a proportional band of 40% to 50% is
suitable.
Let us consider now that the design primary H corresponds to 120 kPa and the required DpL is 30 kPa. If
the primary H increases from 120 to 220 kPa, the secondary differential pressure will increase from 30 to
33 kPa (+ 10%). The disturbance of 100 kPa has been reduced to 3 kPa on the secondary side. Without
the p controller, the circuit would experience an increase of differential pressure of 100 kPa. In this case,
the STAP has reduced the effect of the disturbance by a factor of 33.

Fig 3- Influence of primary differential pressure on the secondary DpL.

In practice, the adjustment of the set point is made, in design conditions, to obtain the design flow that
can be measured with the measuring valve STAD/M. Thats all. There is no need to calculate anything. If
this setting is made when the primary H is higher than its design value, the set point obtained is a little
lower than normally necessary. The difference is generally negligible but can be corrected when the flows
are measured for the balancing report.
The set point also corresponds to a certain number of turns of the setting. The relation between the set
point and the number of turns is given in technical specifications. This information is useful for faster
presetting of the STAP at its required set point, when a measuring valve has not been installed, or when
the present differential pressure H is lower than Hmin.

Art 2-2002

Version 5 -9/09/02

3/6

2- Some applications in air conditioning

2.1 One STAP on each riser

STAD/M

STAP

Fig 4: A p controller STAP stabilises the differential pressure on each riser.

For large systems, the pump head may be too high or variable for some terminals. In this case, differential
pressure is stabilised at the bottom of each riser, at a suitable value, with a STAP differential pressure
control valve.
Note: In heating applications, if all the control valves of one riser shut, the differential control valve
STAP will also shut. All the return piping of this riser decreases thus in static pressure as the water cools
down in a closed area. Differential pressure across the control valves becomes much higher. As a
consequence, the control valve that reopens first will be temporarily very noisy. A minimum flow created
by a relief valve BPV, for instance, avoids such a problem.

2.2 One STAP on each branch

STAD/M

STAP

Fig 5: A p controller STAP stabilises the differential pressure on each branch.

Art 2-2002

Version 5 -9/09/02

4/6

Technically, this solution is better than one STAP for each riser, because the differential pressure
required on each branch may vary. Moreover changes in differential pressures due to the variable pressure
drops in pipes of the risers are automatically compensated.
Regarding the minimum flow, note given under section 2.1 for risers may be extended to branches.
Examples
In figure 6a, each terminal unit C is provided with a balancing valve (STAD) or terminal valve (TBV).
This is the general case examined in figure 5.
In figure 6b, each terminal C is provided with a regulating valve (Trim valve or STK). Since they do
not allow flow measurement in the terminal units, the presettings of the regulating valves have to be
calculated.
In figure 6c, the terminal units are controlled by on-off acting valves with presetting (TBV-C).

Fig 6: One STAP stabilises the differential pressure applied across a set of terminal units.

2.3 One STAP on each control valve

Depending on the design of the plant, differential pressure available on some circuits can vary
dramatically with the load. In this case, to obtain and maintain the correct control valve characteristics,
differential pressure across the control valves can be stabilised with a p controller as represented on
figure 7.

Fig 7. A p controller stabilises the differential pressure across the control valve.

Notes
1- Flow is measured with the measuring valve STAD/M (STAD or STAM), which is an essential
diagnostic tool.
2- When no measuring unit is required (not recommended), the measuring valve can just be replaced by a
pressure connection. In this case the setting of the STAP is calculated based on the Kvs of the control

Art 2-2002

Version 5 -9/09/02

5/6

The given values are: the design flow q and the Kvs of the control valve, which is normally known to an
accuracy of 15%.
The theoretical p to be stabilised by the STAP is shown by the following formula:
q 2
) (kPa - l / h)
Kvs
q 2
p = (36
)
(kPa - l / s)
Kvs
p = (0. 01

The control valve V is never oversized as the design flow is always obtained for the valve fully open. The
control valve authority is and remains above 0.7.
As the secondary DpL is practically constant, all additional primary differential pressure is taken in the
STAP. The control of differential pressure is quite easy in comparison with temperature control if a
sufficient proportional band has been adopted to avoid hunting.

Balancing procedure fig 7

1- Open the control valve V fully.
2- Preset the STAD/M to obtain at least 3 kPa for design flow.
3- Adjust the set point of the differential pressure controller STAP to obtain design flow.
As the flows are correct at each terminal, no other balancing procedure is required.
If all control valves are combined with STAP, then balancing valves in branches and risers are not
necessary unless for diagnostic purposes.

Sizing of the control valve.

Sizing the control valve "V" is not critical in this case. It is however recommended to adopt a pressure
drop at least equal to 20 kPa. As we no longer have to follow the rule pV > 0.25 pump head, a smaller
pV than usual may be adopted, reducing the necessary pump head.

Example with a control valve in injection

Some secondary distributions work with constant flow and variable supply water temperature. A constant
water flow is also required for preheating coils to ensure better freezing protection. For better temperature
control, a constant flow in a unit maintains turbulent conditions and consequently a constant exchange
coefficient. In these cases, a variable supply water temperature is normally obtained with a three-way
mixing valve.
When the distribution is active (with a primary pump), a three-way mixing valve is not permitted as the
flow can reverse in its bypass, due to the primary differential pressure. When the flow reverses in the
bypass of the three-way valve, the mixing function is destroyed. In this case the best solution is to install a
two-way control valve mounted in injection as shown in figure 8a.

Art 2-2002

Version 5 -9/09/02

6/6

If the primary H varies radically, the authority of the two-way control valve will decrease dramatically,
compromising the stability of the control loop. In this case, the best solution is to stabilise differential
pressure pCD across the control valve with a p controller (Fig 8b). If a minimum flow must be
generated to protect the primary pump, it can be created by a balancing valve situated between C and D
(Fig 8c). The adjustment of this minimum flow is quite easy as the pCD is stabilised. Just open this
balancing valve to obtain the minimum required water flow. This minimum flow can also be obtained
with a relief valve BPV. This solution is better than the use of a manual valve because the minimum
flow is generated only when necessary. This reduces the primary flow and therefore the pumping costs.
Some designers prescribe a non-return valve in the pipe AB to avoid any flow going from B to A.
There are two main reasons:
1. For a preheating coil submitted to a low air temperature, the non-return valve allows the
primary pump to inject hot water in the coil if the secondary pump fails. This ensures a
protection against freezing.
2. In district heating distribution, if the two-way control valve is oversized or if the secondary
flow is variable, the water flow in the bypass AB may reverse, reheating the returns. The nonreturn valve avoids this reverse flow.
When such a non-return valve closes, the pressure in B increases, reducing differential pressure across the
control valve, improving its authority.
Conclusion
Particularly to reduce pumping costs and compatibility problems at interfaces, variable flow distributions
are used more extensively than in the past. The main problem associated with this distribution is the
variable differential pressure obtained on the circuits and particularly across the modulating control
valves. Modulating stable control is therefore more difficult to obtain. Variable speed pumps can help, but
in a distribution, just one pressure somewhere can be controlled. One reasonable solution is to stabilise
the differential pressures locally with differential pressure controllers STAP with eventually the help of
variable speed pumps.
With differential pressure controllers STAP: modulating control valves work as required, noise is
reduced and balancing procedure is simplified.
Ref: Handbook 4 - Hydronic balancing with differential pressure controllers Tour&Andersson