Filtration+Separation July/August 2009

Air filtration and efficiency:

Air filters – energy rating and labelling T

he request for reducing CO2 emissions comes from several directions and is becoming part of the marketing strategy of various industries. Energy consumption awareness is increasing, and the need to reduce energy consumption drives manufacturers to introduce energy efficiency indicators in almost every product and engineering field. The air cleaning industry is no exception.

Energy efficiency labels and indicators

In the European Union the practice of energy labelling of home appliances started several years ago and has enjoyed good success and overall acceptance by interested parties (Figure 1). The idea to extend energy labelling to most, if not all, heating, ventilation and air-conditioning equipment should be equally obvious. However, in the case of air filters it is not easy or straightforward to choose the indicator to be used for rating their energy performance. Why is it difficult in this case to choose an “energy efficiency” indicator? To understand the choices to be made in setting an energy rating system for air filters it is necessary to make clear what is behind each option.

The energy efficiency of an entity (a device, component, or system) is commonly defined as the “useful” output divided by the total input. Several examples adopting this approach are present in our everyday life.

For instance the energy conversion efficiency can be defined as the ratio between the useful output of an energy conversion machine or process and the input necessary to obtain such useful output. Since they are both expressed in energy units the resulting efficiency is dimensionless and its value is always lower than 1 (or 100%). The useful

output may be electric power, mechanical work, or heat. Energy conversion efficiency is not defined uniquely, but it depends on the thermodynamic value of the input. All or part of the heat produced from burning a fuel may become rejected waste heat if, for example, work is the desired output from a thermodynamic cycle. Therefore the energy

spent can be weighted with the Carnot factor to take into account that it is not possible to convert all thermal energy into mechanical work.

Another common example is given by incandescent light bulbs: to produce visible light they emit as heat approximately 90% of the power consumed. Artificial light sources

Any new energy labelling or efficiency standard relating to air filters would have to take soot nanoparticles into consideration, which are present in the atmosphere and strongly affect the behaviour of air filters.

are usually evaluated in terms of overall luminous efficacy, i.e. the ratio between the total luminous flux emitted and the total amount of input power. Luminous efficacy of a light source is a ratio of the visible light energy emitted (the luminous flux) to the total power input to the lamp. It is measured in lumens per watt (lm/W) and its maximum value is 683 lm/W (at the peak sensitivity for the human eye).

Even though the above definitions include the notion of usefulness, efficiency is considered a technical or physical term. Goal and mission oriented terms comprise effectiveness and efficacy.

It must be pointed out that the “energy efficiency” notion applied to air cleaning devices and filter media is a fairly recent concept. In fact “filtration efficiency” (which is the fraction or percentage of a challenge contaminant that is removed by a filter at a given time) has been in use for many years and is the primary reason for spending energy to operate an air cleaning device. This use is so widespread that in this article, “efficiency” without a modifier will always mean contaminant-removal efficiency. A less confusing term which characterises particle-removal ability of filters is penetration, [100 – efficiency] in percent, but here we will employ the more generally used term, efficiency.

The importance of efficiency explains why during the last decades so much effort has been spent to prepare test methods and rating systems based only on the efficiency of air filters. Interested parties, like legislators, end users and designers, are now paying more attention to the energy performance as well, and this is driving the filtration industry to take this new requirement into consideration. However, to establish an energy rating system for air filters some difficulties (not trivial at all) need to be overcome.

Parameters to be set in advance

Ideally the indicator chosen for the energy labelling should take into consideration all three parameters that characterise a filter element:

• Particle removal efficiency;

• Air flow resistance;

• Expected service life.

The energy efficiency of an air cleaning process is strictly related to its energy consumption (taking for granted that a certain removal efficiency shall be provided). Energy consumption is the energy used by the fan to overcome the resistance caused by the air cleaning device. A rating system should provide in advance, in some way, how many joules or kilowatt-hours are required to operate an air cleaning device for a certain operating life. Such information is not easy to obtain because the energy consumption of a filter does not depend on its air flow resistance

only. In fact, it is also strongly influenced by other parameters like the actual air flow rate flowing through the system, the fan pressure-vs-flow and horsepower-vs.-flow relationships, the tightness of system ducts, the control logic of the system (variable or fixed volume), etc. This applies also to common household appliances. For example the actual energy consumption of a refrigerator is strongly influenced by the average temperature of the environment in which it is working.

At any rate, filters with the lowest air flow resistance are desirable, even if energy consumption and savings cannot be quantified exactly. An energy rating system should certainly take this into account.

In fact there is no doubt that any filtration system is expected to deliver a certain amount of cleaner air, no matter what its energy consumption will be. In the unlikely case for which the efficiency is not critical (i.e. no minimum efficiency is required), the energy consumption could be minimised by simply removing the air filter from the system.

Obviously the level of expected cleanliness of the air will not be the same for any filter. Fibrous filtration is fractional, i.e. efficiency depends strongly on particle size and nobody would think to use a HEPA filter as the first

filtration stage or to use a pre-filter having low efficiency to clean the air delivered to a clean room. Moreover, efficiency may increase, remain stable of even decrease (in case some electrostatic effects are present, or captured particles are dislodged from the filter media) during the filter service life. This explains the presence of so many different products on the market.

Hence, in defining an energy labelling standard, first of all it is necessary to choose the application, the purpose for which the filter will be used. It makes sense to limit the comparison of the filters’ energy consumption to filters serving the same application or environment. This means that the first item of an energy efficiency rating system must be:

1. Only filters having essentially the same minimum efficiency can be compared to choose which is the most energy efficient. This minimum efficiency may occur at the beginning of filter service lives or after some time in case of media carrying electrostatic charges.

1. The efficiency and the air flow resistance of the filter cannot be correlated. In fact the resistance of common fibrous filters for removing particulate matter is due to three factors:

a) The resistance of the filter medium itself (flat, pleated, mini-pleated or in pockets);

b) For pleated and pocket-type filters, the pressure losses in the airflow paths into and out of the filter media.

c) The resistance of the structure supporting the filtering elements (box, V-shapes screens, etc.)

The first term will change during the service life while the second one will remain constant. There is a wide variety of filter media available on the market and they may supply the same efficiencies with very different air flow resistances. On the other hand, even using the same filter media, two air filters may have different air flow resistances. In fact the amount of media (i.e. the net filtering area) and the way in which it is extended and employed can be drastically different.

There are other constraints inherent in an energy efficiency labelling system:

2. An energy efficiency rating cannot take into account system effects or any other energy impacts external to the filter itself. For example, different fan drives have different energy-consumption characteristics, but drive choice is not related to filter characteristics. The filter’s air flow resistance as a function of air flow rate is the only parameter that defines its energy consumption;

3. Constraint 1 dictates that standardised conditions for rating the energy

Figure 1: The familiar energy label for household appliances.

Filtration+Separation July/August 2009

consumption of the air cleaning device are necessary, not just standardised conditions for rating resistance, as is done in existing standardised filter tests;

4. In order to set a meaningful rating system it is necessary to know how the air flow resistance will change during the service life of the air cleaning device. The results could be misleading by taking into consideration the air flow resistance in clean conditions only;

5. The air flow rate flowing through the air cleaning devices to be rated must be set and be the same for all filters. Otherwise the differences would not be due to the properties of the filters themselves.

Energy consumption indicators

The best filter is the one offering the highest collection efficiency with the least pressure drop. Air filtration scientists have been studying for a long time a way to combine the efficiency and the pressure drop caused by different types of filter media and filters with different construction. Rigorous expressions usually refer to filter media but can be used also for the filters available on the market.

To compare different types of filters and filters of thickness Hinds (1999) and Brown (1993)

suggest the “filter quality” parameter, having the following expression:

pP

qF ∆=

)/1ln( =

pP

∆− )ln(

Where P is the penetration of a mono-disperse aerosol and Dp is the pressure drop. Comparisons of qF must be made for the same filter media velocity and test aerosol particle size.

Other authors had proposed earlier very similar expressions for a “quality factor”. See Dorman (1966, 1974):

pP

Q f ∆−=

log100

These expressions are derived by the classical theory of depth filtration in fibrous filters, assuming that the air flow resistance is proportional to the thickness of the filter medium while the penetration decreases exponentially along its thickness.

The previous expressions are useful but do have some limitations which make them not suitable for a filter energy rating system. In fact these indicators are measured in pascal-1; this is not in line with the expectations of the market willing an indicator expressed in joules or kilowatt-hours that will allow both

engineering and financial analyses. Moreover the quality factor refers to the clean (initial) state, while both air flow resistance and penetration will change during the expected service life of the filter.

To overcome these problems Podgorski (2005) proposed the “filter utility factor” (FUF) which has a rather complicated expression and is a function of the time t. FUF takes into account the variation of the pressure drop and penetration during the filter service life tF exposed to the upstream aerosol concentration c0:

Where

Dp0 is the initial air flow resistance;

uPE is the unit price of energy;

tC is the time constant for a filter (each filter has its own time constant function of several parameters);

h is the efficiency of the fan (which, in accordance with our constraint (2) above could simply be dropped from the definition of this factor).

The FUF is interesting because it takes into account the behaviour of the filter during its whole service life. This is an essential condition to set an energy rating system. However, in order to get useful indications from this parameter it would be necessary to use a synthetic dust capable of a reasonable simulation of the evolution of pressure drop and penetration vs. time in a typical application. Standardised synthetic dusts commonly used today (ASHRAE and ISO “fine”) do not fulfil this requirement (Hanley, 2003). The particle size distribution of the synthetic dusts and the concentration resulting from the dispersing process are completely different from the vast majority of common applications. Therefore the dust loading kinetics currently obtained during laboratory tests do not reflect the in-service behaviour. As a practical consequence the energy rating system based on the FUF obtained using the current synthetic dusts would be misleading or useless. However, this is a limit of the current natural ageing simulation, not of the indicator itself.

A much simpler indicator has been recently proposed by Ginestet (2009) who uses the coefficient K, representing the increase of pressure drop of the filter per gram of dust retained:

DHCp

K average∆=

Where DHC is the “dust holding capacity”, i.e. the amount of a synthetic dust captured by

Without a better simulation procedure, any energy rating system or scheme relating to air filters would probably be of little use. However, there is a strong pressure coming from the market to set up and standardise such a system.

the filter by loading it in laboratory till a fixed final pressure drop (e.g. 300 Pa);

Dpaverage is the average pressure drop during the loading phase in laboratory to reach the same final pressure drop as above.

This indicator must be applied to filters providing the same efficiency, even if the penetration is not included in its definition. The “K” indicator, even if less accurate than the FUF, suffers from the same problem, i.e. the need of a new artificial ageing process providing much better simulation of the variation of the pressure drop and penetration vs. time.

The need for better simulation of air filters’ performance

It must be recognised that, probably, bench tests will never be able to reproduce exactly what happens during the actual service of air filters. This has not been a dramatic problem in the past because the intention was to compare different products in the same artificial conditions to understand which one behaved best. In the case of the energy rating and labelling system this is not acceptable any more. In fact judgments will be made on the energy consumed, so that comparisons based on the overall cost (money or life-cycle CO2 emissions) of

different options (filter A costs X more than filter B but allows Y savings) can indicate the best choice.

The consequence is that bench tests and especially the aerosol used for ageing the filter in laboratory must be improved to provide better tools for the future energy rating systems.

A study was carried out to develop a new loading dust (Hanley, 2003). However, the focus was in reproducing the drop off in efficiency affecting certain types of filters. Some studies are currently being finalised by some other research institutes in Europe and USA. To reproduce the loading behaviour it looks essential to include soot nano-particles that are present in the atmosphere and that strongly affect the behaviour of air filters.

Without a better simulation procedure any energy rating system or scheme would probably be of little use. However, there is a strong pressure coming from the market to set up and standardise such a system. Unless a strong action from all interested parties takes place, circumstances will force air filtration experts to fall into the familiar rhythm of exchanging verifying the reliability of testing procedures for developing ratings that fulfil designers’ actual needs. •

Contact:Paolo Tronville, Ph.D.Politecnico di Torino, Turin, ItalyE-mail: paolo.tronville@polito.itTel: +39 011 0904477

References

(1) Brown R. C., Air Filtration, Pergamon Press, Oxford, 1993.

(2) Davies C. N., Air filtration, Academic Press, London, 1973.

(3) Dorman R. G, In Aerosol Science (ed. C.N. Davies), Academic Press, London, 1966.

(4) Dorman R. G., Dust control and air cleaning, Pergamon Press, Oxford, 1974.

(5) Ginestet A., Private communication, 2009.

(6) Hanley J. T., Owen M. K., Final Report on ASHRAE RP-1190 “Develop a New Loading Dust and Dust Loading Procedures for the ASHRAE Filter Test Standards 52.1 and 52.2”, 2003.

(7) Hinds W. C., Aerosol Technology: Properties, Behavior and Measurement of Airborne Particles, John Wiley & Sons, New York, 1999

(8) Podgórski A., Balazy A., Gradon L, “Optimization issues of aerosol filtration in fibrous filters”, Filtech 2005, October 2005, Vol. II, p.226-233.