Selection of Control Valves
By Chris Parsloe:
Modulating control valve selection for different building services applications can be problematic since it falls awkwardly between designers, installers and equipment suppliers. The new BSRIA guide BG 51/2014 Selection of Control Valves in Variable Flow Systems describes the valve options available and how to make appropriate selections.
Flow control is important because the better control we can achieve over heat transfer, the better control we will have over system operating temperatures. Poor control of heating or cooling outputs translates to larger swings in space temperatures. Furthermore, (and often overlooked) poor control will result in uncontrolled return water temperatures to central plant. If we are trying to maximise the system temperature differential in order to get the best performance from boilers or chillers, then a control valve that does not control is a problem.
In order for any valve to provide good modulating control, its characteristic must match that of the heat transfer that it is trying to control. For example, if there is a proportional relationship between flow rate and heat transfer then the valve’s characteristic should also be proportional (or linear) i.e. the change in flow rate through the valve for a given degree of closure should be the same as the degree of closure. A proportional characteristic is suitable for water to water heat exchangers such as plate heat exchangers.
For forced convection water to air heat transfer, such as fan coil units, active beams and air handling units, a large reduction in flow rate is required in order to achieve a small reduction in heat transfer. Hence, for these applications, a linear characteristic is not ideal. Instead an equal percentage characteristic is best. These valves are so-called because the percentage change in flow rate is always the same for a given percentage change in valve opening.
A valve’s characteristic is influenced by a combination of the internal valve seat and plug arrangement combined with the electronic actuator that is fitted. It is essential that the correct valve and actuator combination is selected in order to achieve the correct characteristic.
However, the valve supplier cannot always ensure that the valve will deliver the characteristic intended. This depends on where the valve is located in the system and the relative pressure losses across the circuits in which it is controlling flow. This introduces the tricky issue of valve authority. In simple terms, a valve and actuator combination might be designed to give you the specified characteristic but if you install it in a circuit for which the resistance across the valve is small relative to the rest of the circuit, then it will not function as intended. Some degree of the valve’s closure will be used up just trying to catch up with the other resistances in the circuit. Only when the valve is the biggest resistance in the circuit will it start to influence flow.
Valve authority is therefore defined as the pressure loss across the fully open control valve relative to pressure losses across the other components in the circuit. Conventional wisdom is that the control valve should be at least equal to the pressure losses in the rest of the controlled circuit giving a minimum authority of 0.5 i.e. the loss across the valve divided by the losses across valve and other components including the valve itself.
Hence, control valve selections are made on this basis. For three and four port valves where the valve diverts flow through a by-pass, valve selection is relatively straight forward. The circuit in which flow is controlled is that through the coil itself so the only pressure loss is the coil pressure loss. This means that the fully open valve pressure loss only needs to match the pressure loss through the coil it controls in order to have good authority. For these valves the coil manufacturer can easily select the valves.
However, two port control valves are more of a challenge since the controlled circuit comprises all pipes and components back to the nearest fixed pressure point this being either the pump of an appropriately located DPCV. To match all of these resistances, two port valves also have to have higher resistances. It also means that the coil manufacturer cannot size the valve. The only party that has all of the appropriate pressure loss information is the designer. But sizing the valves is complicated and best suited to a computerised calculation.
Pressure independent control valves have provided a partial solution to valve selection since they combine a two port control valve with a DPCV. Hence the pressure loss across the control valve is fixed giving it the theoretical authority of one. However, PICVs have their limitations in terms of limiting minimum flow rates and control characteristics. Ultra low flow rates can be a problem and the achievement of an equal percentage characteristic can be a challenge. Furthermore, flow repeatability issues are more noticeable as the valves reply on the flexing of a spring.
The new BSRIA provides some understanding on all of these issues and should be helpful to all engineers involved in valve selection.
Chris Parsloe (Parsloe Consulting Ltd) is the author of CIBSE Commissioning Code W 2010, BSRIA BG2/2010 Commissioning Water Systems, and BSRIA BG 29/2012 Pre-commission Cleaning of Pipework Systems.