Many users are unaware that, during traditional manifold sampling, the particle counter undercounts the ambient particle concentration. Because of the additional vacuum created by the manifold pump, the particle concentration is substantially lower in the air sampled than in the ambient cleanroom air. Unless the system adjusts the data appropriately, this lower-than-expected particle density causes a 7-17% under-reporting of particle concentrations.

Historically, different particle counters have used different methods of compensating for this undercounting. To date, none of these compensation techniques have resulted in truly accurate numbers; some have lead to additional errors, such as incorrect particle sizing, which in turn can cause as much as a 40% underreporting in the first channel.

Particle Measuring Systems has developed a system where these counting errors are automatically eliminated. When used with the Aerosol Manifold II, the LASAIR® II particle counter's proprietary system for pressure monitoring and flow control makes it possible for the first time to correctly normalize the raw counts and calculate accurate cleanroom particle concentrations.

Introduction

The LASAIR II particle counter is the latest generation of aerosol particle counter from Particle Measuring Systems. The Aerosol Manifold II (AM2) is designed to work with the Lasair II, under the control of facility monitoring software (Pharmaceutical Net or Facility Net). This paper reviews the measurement errors inherent in manifold monitoring, then explains the patented pressure measurement and normalization techniques that PMS uses to eliminate these problems.

Particle Undercounting in Manifold Systems

In traditional manifold systems a powerful pump pulls air simultaneously from as many as 32 different monitoring locations; each location can be as far as 125' away. This large cumulative length of tubing, plus the volume of air being moved, requires the pump to create a strong vacuum in the manifold chamber. Thus, when the particle counter in turn samples air from the manifold, it also must operate at a vacuum; the particle counter no longer draws air at room pressure (as it would if no manifold were being used).

This increased vacuum (i.e., lower pressure) in the particle counter causes an error in the measurement of ambient particle concentration. The standard particle counter is designed to generate a 1.0 CFM flow at a constant velocity at the sample inlet pressure. In standalone (non-manifold) use, the sample inlet pressure is essentially the same as the atmospheric pressure, so the unit maintains a 1.0 CFM flow, plus or minus its nominal tolerance (e.g., 5%).

However, with a manifold attached, the entire system operates at a lower than ambient pressure; this is due to the vacuum required to pull air through the long lengths of manifold tubing. As a result, the 1.0 cubic foot sampled at the inlet will have a lower concentration of particles than observed in 1.0 cubic foot of ambient air (see Figure 1). Thus, the particle count per cubic foot with a manifold will be proportionately lower than if sampled directly from the ambient air.

To provide truly accurate results, the manifold system must compensate for this change in pressure. To understand how the Lasair II particle counter's compensation system works, it is helpful first to review its pressure measurement system.

Pressure Measurement in the LASAIR II Particle Counter

The LASAIR® II particle counter uses a new, patented pressure measurement system to avoid the undercounting problems traditionally associated with the use of a manifold. At the heart of the system are three separate pressure transducers:

1. The first pressure transducer monitors the pressure inside the sample inlet. This sensor allows the Lasair II particle counter to measure and correct for the vacuum caused by the manifold pump. (See Figure 2 in pdf file.)
2. A differential pressure sensor measures the difference between the pressures before and after the sample cell. The Lasair II particle counter uses this information to regulate the flow rate and velocity through the sample chamber.
3. A third pressure transducer measures the ambient atmospheric pressure.

Particle Concentration Correction

Airflow in the Lasair II particle counter is servo-controlled using the data from the sample inlet and the differential pressure sensors. This information is used to adjust the Lasair II Particle counter blower/pump to generate a 1.0 CFM flow rate at a constant velocity at the sample inlet pressure.

In normal sampling, this is approximately the same pressure as in the cleanroom. In manifold mode, however, the Lasair II particle counter maintains the same velocity while sampling 1.0 CFM at manifold pressure (i.e., lower than ambient).

Pharmaceutical Net then corrects for this phenomenon. First, it calculates the equivalent ambient volume sampled (instead of the nominal 1.0 CF). This is done by multiplying the measured sample volume times the ratio of the density of the air at the inlet to the density of ambient cleanroom air.

The key formula is:
(1) Equivalent ambient sample volume (CF) = Measured volume x Inlet density / Ambient density,

which can be simplified to:
(1a) Equivalent Ambient Volume (CF) = Measured volume x Inlet pressure / Ambient pressure.

Example: Calculating Equivalent Ambient Volume
Background: Manifolds draw about 25" to 50" H2O (or 6.2 to 12.4kPa) of vacuum. Atmosphere at sea level is approximately 101.3kPa.

1. What is the equivalent ambient sample volume at 25" H2O of vacuum (at sea level)?
   1.0CF x (101.3kPa - 6.2kPa) / 101.3kPa = 0.939CF
2. What is the equivalent ambient sample volume at 50" H2O of vacuum (at sea level)?
   1.0 CF x (101.3kPa - 12.4kPa) / 101.3kPa = 0.878 CF

Normalizing Counts in Pharmaceutical Net (or Facility Net)

Once the equivalent ambient sample volume has been calculated, the raw particle counts can be correctly normalized:

(2) Normalized particle count (N/CF) = Observed particle count / Equivalent ambient sample volume,

which is equivalent to:
(2a) Normalized particle count (N/CF) = Observed particle count x Ambient pressure / Inlet pressure.

In the Lasair II particle counter manifold system, both raw counts and equivalent ambient sample volumes are communicated to Pharmaceutical Net (or Facility Net), which then performs the normalization calculations. (For example, for an equivalent sample volume of 0.939, the normalization factor would be 1.0 / 0.939 = 1.065.)

Correction/Normalization Table

Table 1 provides the equivalent ambient sample volumes (compared to the nominal 1.0 CF sample volume) for a wide range of manifold operating conditions. Notice that the equivalent sample volume may need correction for altitude as well as for manifold vacuum.

As can be seen when using a manifold system, the equivalent ambient volume sampled typically will be only 83-93% of the 1.0 CF that the particle counter sampled. Systems that fail to compensate will under-count particle concentrations by 7-17%.

LASAIR II Particle Counter Flow Rate

To ensure that the user has full information for accurately normalizing the particle concentration, the Lasair II particle counter displays the equivalent ambient sample flow rate (instead of the actually sampled flow rate).

When using a Lasair II particle counter with the AM2, it is normal for the flow rate displayed on the L2 to be more than 5% lower than 1.0 CFM without triggering any alarms. In fact, a flow rate as low as 0.85 CFM could be validly displayed.

To verify that the Lasair II particle counter is operating at the correct flow rate, use a pressure monitor to measure the pressure at the L2 sample inlet, determine the approximate altitude, then compare the flow rate displayed on the L2 to the equivalent flow rate given in Table 1; the two numbers should be within 5% of one another.

Shortcomings of Alternative Compensation Techniques

Over the years, particle counter manufacturers have employed a variety of techniques for compensating for this undercounting phenomenon. The most common has been, upon installation, to increase the pump speed so the equivalent ambient volume is closer to the nominal 1.0 CFM.

As another alternative, some have used a mass flow controller to servo-control the pump. Such controllers automatically increase the volume flow to ensure that the particle counter samples the target mass. Thus, the same mass is sampled as in 1.0 CFM of ambient air.

Unfortunately, both of these techniques require increasing the velocity through the sample chamber. This causes two problems: First, increasing the velocity causes the traditional particle counter to lose its calibration, to the point that it no longer meets JIS specifications.

Second, increasing the sample velocity also causes a new type of problem: particle-sizing errors. Counters determine particle size by evaluating the shape of the pulse a particle generates as it passes through the laser beam; the shape recognition algorithms used to accomplish this are set during calibration. If the same particle is passed through the beam more quickly, the pulse is smaller; this results in some particles being misclassified as belonging to a smaller size channel. In the smallest size channel, some particles will not be counted at all.

Thus, increasing the sample velocity reduces the counting efficiency. Moreover, where before one could calculate the degree of undercounting (see Sections 4 & 5), now the size of the error becomes very complicated to predict.

Figure 3 illustrates particle-sizing errors caused by increasing the sample velocity. This graph shows three separate tests run on a 1.0 CFM particle counter connected to a manifold via a 125' sample line. The first bar illustrates counting losses (in 0.3 µm particles at 50" H2O pressure) when the sample velocity was increased to achieve an equivalent ambient flow rate of 1.0 CFM. Note that the counts dropped to under 60% of those seen by the reference unit. While this data reflects the first channel, the higher channels also were impacted.

In the next two bars on Figure 3, normal sample velocity was maintained so that the unit drew its normal 1.0 CFM at constant velocity at the sample inlet pressure. For these cases the normal counting efficiency was maintained; counts were approximately 100% of those seen by the reference unit.

The Lasair II particle counter avoids these undersizing errors, as well as the original undercounting errors. It avoids undersizing errors by maintaining the same sample velocity throughout. It then uses Pharmaceutical Net (or Facility Net) to correctly normalize the data to eliminate undercounting errors (see Sections 2.0, 5.0, and 6.0).

Conclusions

The powerful vacuum of the manifold pump causes the particle concentration of the sample to be lower than the particle concentration of ambient cleanroom air. Unless the system compensates properly, this lower concentration results in a corresponding under-reporting, typically under-counting by 8-20%.

Previous manifold sampling systems have either ignored this error, or have used some system of compensation. Until now, none of these compensation techniques have produced truly accurate counts. In part, this is because most have been based upon increasing the sample velocity; this in turn has generated particle under-sizing errors, which also resulted in as much as 40% under-reporting in the first channel.

The Lasair II particle counter offers the only manifold system that corrects for both of these issues. Using data from the patented pressure-sensor system, the L2 calculates the pressure-corrected, equivalent ambient sample volumes, then sends Pharmaceutical Net (or Facility Net) the normalization factors required to produce accurate measurements.

Contact us if you need more information or have questions.

LASAIR® is a registered trademark of Particle Measuring Systems, Inc.