The laws of physics though, stay the same. When it comes to designing energy efficient mechanical ventilation, no amount of clever electronics or high efficiency heat recovery can compensate for high pressure drops, wildly extended hours of operation, or excessive air change rates. If a system has any of these vices, your carefully modelled energy targets will miss by a mile.
Specific fan power is the conventional way of describing the total electrical drive power capacity of a ventilation system, including both supply and exhaust fans. But how do you design a mechanical ventilation system with low specific fan power? Can you anticipate performance penalties caused by high pressure drops, variable occupancy patterns and specification busting by the contractor? Are inverter drives a panacea, or is two-speed control a better bet?
Choosing the right fan
Fans fall into two basic types: the popular centrifugal fan in forward and backward curved configuration, and in-line axial fans. Simple forward curved centrifugal fans are cheap, say around £140 for a 2-3 m3/s unit, but only around 60% efficient. Backward curved centrifugal fans are 80% efficient, but about three times more expensive.
Axial flow fans are not as efficient as backward flow fans, at around 70-75% efficiency, but no belts or pulleys are required and the direct motor can be controlled via an inverter and/or variable pitch blades. They are not suitable for small volumes. Prices are similar to those of forward-curved centrifugal fans.
The efficiency of a centrifugal fan is dependent on the nature of the motor drive: indirect drives using belts and pulleys can cause performance losses of around 10-15%. Although the popular solution is to use direct drives and speed inverters, performance improvements are not guaranteed.
Having chosen the type of fan, the designer then has to make a fan selection based on an assessment of the loads (including system losses) and the occupancy profile of the space being ventilated. All fan manufacturers provide performance curves for their products which enable designers to choose the appropriate fan for the duty.
However, this is where the problems start. Designers need to understand the precise nature of the pressure drops that the fan will have to overcome, but without being so conservative that the fan ends up operating at the bottom of its fan curve and therefore inefficiently.
Fans that are asked to deliver air at the lowest part of the fan curve become noisy, inefficient and prone to stalling. Adjusting belts and pulleys and reconfiguring inverter drives can solve the problem, but at the penalty of a higher volume than is required, and a lower fan efficiency.
It is worth noting that clients often see higher air change rates as beneficial, but the penalty is high specific fan power, high supply volumes and an inability to match fan operation to demand.
Calculating pressure drops
While designers need to factor in the total resistance to airflow, the air handling unit components like filters, dampers and coils form a substantial part of the overall system resistance. These elements largely dictate the kW rating of the fan.
Table 1 shows typical pressure drops within an ahu. The filter pressure drop should be calculated "on the dirty", which is usually twice the pressure drop of the filter when new.
John Corps, executive sales engineer with Woods Air Movement, advises that the static pressure within an air handling unit can easily be around 1200 Pa, and far higher when elements are added like heat recovery devices and desiccant dehumidification.
Although many of these pressure drops are an inevitable consequence of air treatment, designers need to think very carefully about the energy penalties of installing so-called energy efficiency devices in the airstream. PROBE studies record that rather too much emphasis can be placed on saving heat without considering the effect on C02 emissions.
Electricity contributes 0·52 kg/CO2/kWh compared to just 0·2 kg/CO2/kWh for gas, so it is clear that an heat recovery device would need to save three times a unit cost of fan power to be worthwhile in overall energy terms. Given that the heat recovery will only be required for perhaps 30% of a building's annual running hours, serious consideration should be given to bypassing the pressure drop for the rest of the year.
Clients and quantity surveyors often regard the minor cost of bypass ductwork as an expendable item, therefore sometimes designers' time is better spent on providing intrinsically efficient heat raising plant and rather less worrying about heat recovery. An exception to this rule is the full fresh air system, for which heat recovery is arguably essential.
If a bypass does get past the value engineering exercise, bear in mind that the lower static pressure in the index loop when the bypass is in operation may compromise the control authority of any terminal devices. This usually forces the use of a throttling mechanisms to maintain the desired static pressure.
Inlet guide vanes can be used to maintain pressure, but inverters are more popular. If they are used, bear in mind that the fan must not move too far (or too often) down its curve. The loss of fan efficiency could cancel the benefit of the bypass.
Terminals and index loops
Engineers...put inverter drives on motors to save energy, but then forget about the fan
The total resistance to airflow of a ducted system is the sum of the resistances of the individual duct lengths, fittings and other components like terminal units. Systems are usually balanced using an index circuit2,3.
The index circuit is usually measured on the longest run from the fan to the furthest terminal unit, and the static pressure then used to control the fan speed. This is usually employed for vav systems.
The need to maintain static pressure in the index circuit can have implications for fan energy consumption. Seemingly minor details such as the location of sensors can have a major impact on fan energy. PROBE research has shown that controlling static pressure off an index circuit which is unaware of the action of individual terminal boxes can cause fans to run at full capacity.
If the ventilation system is fully loaded, fine. However, if the demand is led by a single room or even a single terminal unit – a classic "tail wagging the dog" situation – the choice of central plant may need to be questioned.
Static pressure sensors placed at the start of duct systems may not provide the level of control required at a powered terminal unit. One solution is to place the static sensor at the last box, and provide the minimum static pressure for the terminal's proper operation. The damper position in the terminal unit can then be used to control the fan speed.
If maximum damper position is set at, say, 60% rather than 90%, that will reduce static pressure significantly and thereby reduce fan energy consumption – if the fan is sized correctly.
Inverters or two speed control?
The advent of stepless control has been a godsend to designers of all types of pumped circuits which have to accommodate varying loads with a decent design margin. However, inverter drives and eddy current motors are not a surrogate for proper fan specification.
"Many engineers don't understand how fans operate" says John Corps,. "They put inverter drives on the motors to save energy but then forget about the fan. They then wonder why the fan is vibrating. Of course, it's operating too low on its curve." Depending on the loads and level of occupancy, there may be an argument for several smaller air handling units. It may even be more efficient to put axial fans in ductwork legs, or, as did Oscar Faber at the headquarters of Britannic Assurance, to use multi-zone air handling units equipped with individual fan sections1.
Designers should consider the virtues of two-speed control. Two-speed systems make fan selection easier, are more likely to maintain good fan performance efficiencies and are potentially easier to manage. However, they require good interface engineering and careful balancing.
"Two-speed ventilation systems demand a level of human intervention which should be both visible and measurable," says building physicist Bill Bordass. "With variable speed control you often don't know what it is doing. It also tends to fit a perceived demand rather than an actual demand, so if there is a faulty sensor or an incorrect setpoint you can end up delivering more air than you actually need." Bordass believes that a two-speed system that only operates on high speed for night cooling and peak loads can be a good alternative. It could also be made to be a little noisy at the high speed setting to tell occupants when it is running. "This is tricky stuff and designers don't like to do it," admits Bordass, "but it's often useful feedback to the occupier."
Energy analysis and benchmarking
The Energy Assessment and Reporting Methodology is the latest tool for the assessment and reporting of buildings and their energy use. This uses a common language and procedure for estimating the annual energy use of all systems during a building's design, construction and operation.
Each factor in fan energy consumption can be benchmarked against best practice figures from Energy Consumption Guide 19. More important perhaps, it does so in a way that lay clients can understand.
The procedure, now enshrined in CIBSE Technical Memorandum TM22, centres on so-called energy trees. Designers can create branches for heating, cooling, ventilation and lighting, and any sub-sets such as toilet ventilation or car park lighting (figure 1).
The branches of the trees enable all relevant information to be multiplied to arrive at a total annual energy use expressed in kWh/m2/y. For ventilation, these are the supply and exhaust rate in litres/s/m2 (treated floor area), the fan efficiency (expressed in W/litre/s), the hours of use and a management factor. This is the additional fan time the client thinks will be required, for preconditioning, office cleaning or late working.
The ventilation rate is obviously a matter for discussion with the client, but today's good practice is typically 10 litres/s/person. For a typical occupation density of 14 m2/person and a ceiling height of 2·7 m, background ventilation equates to about one air change per hour (or 2 ac/h for minimum fresh air systems). A maximum of 6 ac/h (4·5 litres/s/m2) is typical for all-air office air conditioning5, while night ventilation rates are in the range 2-4 ac/h. In practice these rates vary.
There are several ways of expressing fan efficiency, the most common being W/litre/ s. Generally speaking anything at or below 2 W/litre/s in a mechanically ventilated building can be considered good practice, although some designers are getting down to 1 W/litre/s (table 2). The SCANVAC publication Classified Air Distribution Systems, is a useful document which lays down procedures for designing systems for low specific fan power1 Hours of use is obviously a best estimate by the client, as is the control and management factor. Nevertheless, the virtue of this approach is that everything is visible and everything can be counted. If the kWh/m2 value is excessive, it should be easy to spot the source of the trouble. For example, toilet fans may be considered trivial in energy terms, but not if they run continuously at high speed. Once that is revealed by the energy analysis, the engineer can argue convincingly for demand-controlled toilet ventilation, with proportional control as required.
The key to using an energy tree is to have a good database of information. If you want to add fine detail or coarse detail you can, but as Bill Bordass says, "don't graunch it into the database, tag the data with a simple quality assurance system which says where it came from, and whether it was a guess, an estimate, or a measurement. If it's estimated the designer should say whether it was done to a standard procedure. If it's precise, then describe the context." The energy tree approach allows the designer to demonstrate to the client who owns the problems and who is taking the risk. For example, the penalty of agreeing to the contractors' demand for a cheaper fan. This may become increasingly important when carbon controls kick in. Property developers can then be told the penalties of going small, fast, high pressure and cheap.
Basic rules for low fan power
The following rules should be used to achieve the best possible fan power efficiency:Source
Building Sustainable Design
Postscript
1Classified Air Distribution Systems: Guidelines and Specifications, The Swedish Indoor Climate Institute 1992. 2Bunn R, 'Clean Bold', Building Services Journal, 11/96. 3HVCA Energy Efficiency Handbook ,1987. 4CIBSE TM8: Design Notes for Ductwork. 5CIBSE TM22: Energy Assessment and Reporting Methodology, CIBSE, 1999. 6Parsloe C, 'Rules of Thumb', BSRIA Technical Note TN17/95, 1995. BSJ would like to thank building physicist Bill Bordass and John Corps of Woods Air Movement for their help in researching this article.
