Photo courtesy of Mark Faircloth, Digital Air

Air conditioning is responsible for substantial electricity consumption and peak demand in most of the United States. Over the past decade, energy conservation researchers have studied air conditioning more and more.  Much of this research has focussed on the impact of air flow, duct leakage, refrigerant charge level on cooling performance.

One area which has been neglected by researchers is fouling of evaporator (inside unit) and condenser (outside unit) coils.  Every technician will tell you that every time they look at these coils, they are dirty. Air conditioner capacity and efficiency is a strong function of air handler flow - a fouled coil will allow less air to flow past the coil (not to mention damage to the compressor that can occur with dramatically reduced flows).  A standard part of routine AC maintenance and residential commissioning is to clean the evaporator coil with a wire brush and detergent or other cleaning chemistry and to clean the outdoor coil of leaves and other debris– it is not clear how often technicians really complete this part of the work or whether this really improves the situation.
 
 

Coils can also foul with bioaerosols and other biologically active material.  This can lead to indoor air quality problems.  LBNL report #47669 is about this specific issue.
 
 

Photo courtesy of Mark Faircloth, Digital Air	Photo courtesy of Indoor Atmospheric Science

The Energy Performance of Buildings Group has begun a project to examine the mechanisms that cause evaporator fouling and to try and quantify its impact on cooling performance.  The original intent was to concentrate on residential systems, but some preliminary fieldwork has shown commercial heating and coiling coils have the same geometries and are prone to the same type of fouling and thus they are now included as well.
 

Typical fouling materials include dust and pet hair that bypasses filters, smaller particles from indoor air (i.e. from cooking), and environmental tobacco smoke (ETS).

The project has four components: experiments, fieldwork, modeling, filter bypass, and analysis of performance impacts.  Currently the experiments are well underway and a detailed magnitude analysis has been completed.
 
 

The purpose of the experiments is to determine how and where coils foul.  We had a special evaporator coil manufactured that has the same specifications as a typical residential coil, but is enclosed in a six inch square duct.  The coil is installed in an apparatus that is being used by another researcher to look at particle deposition in commercial ductwork.  Air with particles of a known size are blown into 80 feet of duct work and then, after the flow has had an opportunity to develop fully the concentration of particles is measured upstream and down stream of the evaporator coil. The difference is how much material deposits.

The schematics below show the experimental apparatus.  Liquid particles are generated with a vibrating orifice particle generator.  In the future, this will be expanded to include some runs with solid polystyrene particles.  Electric charges inherent to the particle generation process are then removed with a radioactive source in the charge neutralizer.  The particles are then fed into a mixing box where a fan pushes them into 75' of straight duct.  After the flow has fully developed, the particles pass through an evaporator coil.  Particle concentrations are measured up and downstream of the coil.  Additionally, after each experimental run, the coil is removed from the apparatus and a buffer solution is used to extract the deposited particles on the coil.  This provides an additional confirmation (and is a more accurate measurement) of the particle deposition fraction.
 
 

More recent experiments have included a refinement of the extraction process to allow the separation of the leading edge from the bulk of heat exchanger.  Even more recent experiments have cooled the heat exchanger to measure the impact of thermophoresis (particle motion down a temperature gradient), diffusiophoresis (particle motion due to concentration, in this case, humidity, gradients), and impaction on condensed water on the coil.

An additional experiment using the same apparatus was done to measure coil pressure drop as a function of fouling.  This work was done using SAE coarse dust.

Experimental results appear in all publications.  I am currently working on a journal article which has all of the details of the experiments. Contact me for details.
 
 

In order to generalize the experimental results to other coil geometries, understand the importance of different deposition mechanisms, and predict fouling rates and performance impacts, we created a model.  It takes as major inputs particle diameter, coil fin spacing, and air velocity.  Minor inputs include details about the coil geometry (fin thickness and corrugation, tube diameter),  particle density,  coil temperature,  air temperature and relative humidity, and several other parameters.  It outputs the overall deposition fraction, as well as deposition by each individual mechanisms. Deposition mechanisms currently included are gravitational settling, Brownian diffusion, impaction and interception on fin edges and refrigerant tubes, thermophoresis and diffusiophoresis to fins, turbophoresis, and inlet inertial effects resulting from turbulence in the duct or plenum upstream of the coil.  The model is discussed briefly in LBNL report #47668 in much more detail in #49339 (click here to go to my publications on coil fouling) .
 
 

Over the course of this project, I learned very quickly that almost all coils get fouled and that much of the material that fouls them is relatively large (i.e. dust fibers that are 100 µm - a few cm in length).  These large particles should be stopped by even the coarsest filter.  Since most systems have a filter, it is clear that many filters are not behaving as designed.  There are two possible ways for air to bypass a filter.  The first is leakage into negative pressure (return) ductwork.  Return ducts are often located in dusty buffer spaces like attics, crawlspaces, and dropped ceilings.  Any leaks in this duct work could potentially suck in particles.  For more information on duct leakage, click here for residential duct systems and here for commercial duct systems.  Another, prominent cause of filter bypass is poor installation and maintenance of filters.  The following pictures show some examples of filter bypass.
 
 

A sandwich bag in the fan blades is a good sign that the filter is being bypassed.  The picture on the above right shows that the filter is poorly installed such that much of the air that flows through the System bypasses the filter.  The picture on the right shows an even bigger bag in a AC system.	Photograph courtesy of Mark Faircloth, Digital Air
Photograph courtesy of Paul Francisco, Ecotope Inc.	Photograph courtesy of Paul Francisco, Ecotope Inc.
The picture above is a little hard to fathom until you realize that it is a filter that has been sucked out of its slot and has wrapped itself around an air handler fan.  The picture on the right is a side view of the same system.

Filter Bypass is one reason that coils foul.  FIlter efficiency, deposition in duct work, indoor particle concentrations, and coil design all influence the rate of coil fouling.  I have analyzed each of these factors for residential systems and modeled coil fouling times (the time that it takes the pressure drop to double) for a variety of residential systems.  The short answer is that coils foul in about 10 years, with substantial variation due largely to differences in filter efficiency and indoor particle concentrations.  This results in a performance impacts of 2 - 4 % (capacity and efficiency degredations) for a well designed system, but it can be much larger for marginal systems.  All of the details and more can be found in LBNL #49757 (click here to go to my publications on coil fouling).  The analysis is currently being done for commercial buildings as well.
 

Siegel, Jeffrey, I. Walker & M. Sherman.  2002.  “Dirty Air Conditioners: Energy Implications of Coil Fouling”  Submitted to the 2002 ACEEE Summer Study on Energy Efficiency in Buildings.  Currently available as Lawrence Berkeley National Laboratory LBNL-49757.

Siegel, Jeffrey A. and W.W. Nazaroff. 2002. "Modeling Particle Deposition on HVAC Heat Exchangers."  Submitted to the Indoor Air 2002 conference.
Currently available as Lawrence Berkeley National Laboratory LBNL-49339.

Siegel, Jeffrey and I.S. Walker. 2001.  “Deposition of Biological Aerosols on HVAC Heat Exchangers.” To appear in the proceedings of ASHRAE IAQ 2001. Currently available as Lawrence Berkeley National Laboratory LBNL-47669.

Siegel, Jeffrey & V.P. Carey. 2001.  “Fouling of HVAC Fin and Tube Heat Exchangers.”  To appear in the proceedings of the United Engineering Foundation Conference on Heat Exchanger Fouling 2001.  Currently available as Lawrence Berkeley National Laboratory LBNL-47668.

Additional results and reports on coil fouling will be posted to this page as they are written.  LBNL rules prevent data or reports to be published on the web until they have completed the internal review process.