- 08 May 2006 -
Is it Time To Change Your Filters?
By Rich Thelen, rthelen@globalfinishing.com

What is the best rule for signaling the time to change filters? This question has been asked of spray booth manufacturers, as well as filter manufacturers.
There is not an easy answer. Let's look at filtration, fans, and chamber design.

Filtration
This discussion will concern itself with three-stage NESHAP filters as mandated in CFR 40 part 63. This is the aerospace NESHAP regulation that mandates the use of three-stage filters.

Figure 1: Typical Method 319 testing report. (Published with permission from AJ Dralle Co.)

Please note that the three stages of filtration must meet testing in accordance with Method 319 (Figure 1), which is explained in the regulation (see Tables 3 and 4 of 40 CFR 63.745).

Filter manufacturers have tests that show their filters will perform over a great range of pressures. One manufacturer says that you should change your filters when the first stage is 0.50 in. water static pressure (SP) above the initial pressure drop as measured by a manometer. The second and third stages should be changed when they are at 0.75 in. SP above their initial pressure drop.

Users have different rules. Some users will tell you to change your filters after they are one inch SP above their initial pressure point. Still others have different rules involving filter loading from a visual standpoint. It is even possible to purchase a filter monitor for your filter system that will allow monitoring of the pressure drop across all three stages of filtration. Thus, you will be able to intelligently change just a single stage of filtration, or two stages, or any combination of all three.

The monitor has Magnehelic gauges that display the pressure drop of each stage.

Fans
The answer is more complex than any of these rules. One must look at the entire system and review the fan curves.

Figure 2: Fan curve for Greenback van axial fan.

Figure 2 resembles a typical vane axial fan by Greenheck. In this case, we will use the fan curve for fan #VAB-48F30-I30. We will first describe the various elements of the fan curve. The solid line is the recommended design curve for a particular fan, while the solid line marked "Operating point …. surge point curve" represents the points along the curve at which the fan has a potential to become unstable. Suppose the fan rotates at 979 RPM, then at zero pressure differential (ΔP) the air flow will be 38,000 CFM.

If the fan rotates at 1,142 RPM, then at zero ΔP the air flow will be approximately 45,000 CFM.

Constant RPM: As the ΔP across the filters and the exhaust ducting increases, the air output will decrease. For the fan rotating at 979 RPM and ΔP of 1 in. SP, the air flow will drop from 38,000 CFM. If ΔP increases to 3.05 in. SP, the fan will become unstable and you will not be able to predict the air output. Designers do not want to operate in the unstable range, hence, for this fan and this RPM, you would want to change the filters long before they are at 3 in. SP.

Variable RPM: Spray booths that are fitted with fan controls can automatically increase the RPM when a sensor measures an increased ΔP. As you load the filters with overspray, the RPM will continue to increase, thereby maintaining the same CFM output.

In practice, the maximum ΔP will occur when the filters are loaded to saturation. This is the point at which any additional paint overspray will simply fall to the floor rather than adhere to the filters.

When fan curves are analyzed, the CFM delivered will vary as the pressure drop changes. The fan designer should make an effort to ensure that from the initial starting point ΔP to the maximum anticipated ΔP, the fan should operate as far to the right of the surge (unstable) curve as possible. It is difficult to find a fan that will have stable performance over a static pressure range of two to three inches, and in the example that we will discuss below, you will see why that is so.

Example: Using the same fan curve in Figure 2, let us look at the system characteristics of a typical airplane paint booth. It may require two fans of 35,000 CFM each with an initial pressure drop of 1 in. SP, including filters and duct losses. The design fan curve in Figure 2 shows that, at 1.0 in. SP, the fan is very stable (the point is to the right of the design fan cure.) At 3 in. SP and 35,000 CFM, we are operating directly on the fan curve. At 4.5 in. SP and the same air flow, the operating point falls between the design curve and the surge curve and the fan has a greater potential to be less stable. In the hypothetical case that we were to operate the spray booth so that pressure differential across the filters increased to 7.6 in. SP, at 35,000 CFM, the operating point would fall on the surge (unstable) curve, and we would almost assuredly experience unpredictable variations in velocity.

From Figure 1, we saw that above 3.37 in. SP the filters will no longer hold paint overspray, so operation above this pressure is not required. Ductwork losses may require the fan to operate above this limit, but usually ductwork losses are in the range of 0.25 to 0.50 in. SP, since the exhaust is just ducted immediately outdoors.

Horsepower considerations: The dotted lines correspond to the horsepower (HP) required to drive the fan motor at a specific RPM. We will start with 979 RPM, 35,000 CFM, and 1 in. SP. If we move along the y-axis that corresponds to 35,000 CFM, we see that at 3 in. SP, to maintain this air flow we will need to increase the fan rotation 1,142 RPM, and the motor will need to deliver 27 HP. For the sake of this example, if we were to allow the filters to become excessively blocked with overspray, the ΔP might increase to 4.5 in. To maintain 35,000 CFM, the fan rotation would need to increase to 1,300 RPM and a motor capable of delivering 40 HP would be required.

Figure 3: Magnehelic gauges moniter the pressure differentials across each filter stage.

Effect of current frequency: A variable frequency drive (VFD) changes the frequency of the electric alternating current, and is measured in Hertz (Hz).

Suppose the design engineer selects a 40-HP, 60-Hz motor that can drive the fan at 1,300 RPM against a ΔP of 4.5 in. SP to give an air flow output of 35,000 CFM? Since there is a linear relationship between RPM and Hz, at the initial ΔP of 1 in. SP the motor will only need 979*60/1,300 = 43 Hz, and the costs to operate the motor over eight hours of spray booth operation will be fairly low.

However, as the filters start to load with paint overspray and the ΔP increases to 3 in. SP, the frequency required to drive the motor at 1,142 RPM is 1,142*60/1,300 = 53 Hz. We see, therefore, that as the pressure ΔP increases we need to increase the RPM of the fan, which understandably requires considerably more energy to drive the motor.

When the filters are fresh and new, the speed of the fan is low and the VFD is at a low frequency. As the filters become more dirty, the speed increases and the amperage (power requirement) increases. Finally, we get to a point where the VFD is at its maximum setting and no further speed increase is possible. At this point, the alarms sound and the system may shut down or begin a warning sequence.

What if the fan stayed at a fixed speed or if we did not have a variable speed system? Filters would continue to load and airflow in the booth would be reduced. It is possible to lower the airflow in the booth to levels that are unsafe from a fire risk standpoint.

But, it is more likely that the painting performance of the booth would suffer, causing blemishes in the surface that require rework. The painter would notice this and investigate the cause of the problem. He would soon see that the filter pressure drop is above the set point and that the fan speed is maxed out. He would then deduce that the filters require changing and take appropriate steps to remedy the situation.

Upon examining the state of the filters, he would see that the pressure drops across each stage of filtration are not at their effective limits. In other words, the filters had more life in them. This is not surprising when we look at this from a system approach. The filters are not the bottleneck in the system. The performance of the fan suffers long before the filters have been completely consumed.

The reduction in booth performance would be a vital telltale sign to the painter that he needs to change the filters.

Filter Chamber Design
Now that we have learned about filter and fan characteristics, let us look at the design of filter chambers. Chambers designed for high pressures have several common characteristics. They may contain structural elements, such as I-beams, and they may have heavier sheet metal panel elements. The design of a normal chamber handling 1.25 in. static pressure is a trivial exercise and easy to accomplish.

The design of a chamber for three inch static pressure must have some structural elements, such as I-beams, to prevent the suction from collapsing the walls. In addition, the breaks on the panels themselves must be deeper to avoid the collapse. A designer must also look at the span of the panel and beam elements to assist in the load carrying duties. Obviously, structural design is beyond the scope of this article, so we will go no further into it.
It should be obvious that the three (filters, fans, and chamber design) are interrelated and what affects one affects the others. If the filter changing pressure drop is set at 1.25 in. SP, then the horsepower will be low, the filter replacement cost will be high, and the cost of the chamber will be low. With a chamber designed for 3 in. static pressure, the horsepower will be moderate, while the cost of the chamber will also be moderate. Filter replacement costs will remain moderate, also.

For a chamber of 8 in. static pressure, filter replacement will be low, but horsepower and chamber costs will be extreme. For this reason, prudence suggests an optimal design of about 3 in. static pressure for the fan, chamber and the filter replacement.

Summary
Designing an optimal system of fans, filters, and chambers is a difficult task and requires some trade-offs. But once the trade-offs are analyzed, an economical system is realized. We suggest that filters be replaced at a total filter pressure drop of 3 inch static pressure.

This implies that the chamber is structurally designed to withstand this pressure. The filters will be able to hold all of the paint overspray without making a mess on the floor. The fan will be very stable over its intended range. The filter chamber can be economically designed to withstand the pressures.

Further, with a range as great as this, it is essential that the prudent designer use variable speed fan motors to compensate for the wide range of paint booth air velocities that are possible. The paint booth airflow should be constant in the selected pressure range. Any attempt to operate above this range will result in alarms and attenuation of system performance and an ever noticeable performance drop-off from a painting standpoint.

For more information, contact the author at (e-mail) rthelen@globalfinishing.com.