By Matthew C. Lemleux, The NEBB Professional — Q2 2019 Edition
As presented by the International Standards Organization, cleanroom differential pressures are an indispensable controlled environment necessity. Room pressure differentials are often the first qualification seized upon by regulators and inspectors, and many regulations mandate the installation of visible pressure differential indicators and even continuous monitoring and logging.
Inconsistencies and excursions of room pressure are a continual source of potential headaches for clients. Often, relative room pressurization is neither well understood by the client nor the inspector. Confusion arises generally in distinguishing pressure terminology such as differential, absolute, ambient, relative to, positive and negative. Further, it is important to realize that specified air differential pressures are surprisingly minute. A variation of as little as 0.0012 percent of an atmosphere can warrant an inspector’s or regulatory agency’s attention. Unfortunately, regulators can have an alarmingly Calvinistic interpretation of pressure guidelines.
Despite this criticality, cleanroom engineers often incorporate constant airflow rather than constant pressure control systems into their designs. Regulators are not concerned with the constancy and tolerance of airflow, nor should they be, provided that the airflow at least meets the minimum air exchange rates. How and why this design strategy persists requires a little background investigation.
Cleanrooms are unlike any other type of building construction. A well-constructed cleanroom is more akin to a submarine or spacecraft than to conventional laboratory space. Tremendous effort is expended in sealing and tightening any and all architectural details to minimize the potential for any air leakage. We can use a term like porosity to describe the conductance of air into and out of a particular room.
For example, in a conventional research laboratory, the architectural details are not much more sophisticated than in an office space. Cleanroom construction endeavors are intended to minimize porosity and conductance or, equivalently, maximize resistance to airflow infiltration and exfiltration.
Because of the second-order relationship between airflow and pressure through a gap, a conventional porous space will not register any measurable pressure differential on a gauge despite the net outflow of air relative to surrounding spaces. By contrast, a cleanroom’s resistance causes a measurable pressure differential that can be displayed on an ultra-low-pressure wall-mounted gauge.
Similarly, the same change in airflow produces a much more amplified pressure signal in a cleanroom than in a porous laboratory space. Laboratories may have a requirement to be negatively pressurized to the surrounding occupied spaces due to the nature of materials being handled therein.
In HVAC engineering, this is usually accomplished by continuously monitoring and adjusting laboratory supply and exhaust airflow quantities to maintain a net influx of air into the laboratory from the surrounding space. This deficit or surplus between laboratory supply and exhaust devices is often referred to as offset and is usually on the order of 200 cubic feet per minute (CFM) per laboratory.
When this same strategy is applied to cleanrooms for the purpose of pressure control, a number of unforeseen consequences can arise. Without examining these consequences in detail, they may involve door openings, which have the effect of increasing perceived airflow, rooms being actually pressurized by exfiltration from adjacent rooms rather than their own supply and exhaust terminals, or degradation of non-porous seals, such as door weather-stripping among many others.
By definition, attempting to control room pressures by controlling airflow is what is called an open-loop control system. Controlling room pressure by controlling room pressure is what is called a closed-loop control system. A domestic toaster is an example of an open-loop control system. It purports to control the condition of cooked bread by controlling the time that the bread is exposed to heat. The toaster system (once the timer is set) has no regard for the thickness of the bread, its frozen condition, type of grain, color, or physical appearance. Toast burns because there is no system that monitors its temperature and crispness during the toasting process.
The design used in this facility was to control differential pressures between cleanrooms A, B, C, and D, along with the warehouse space. Upon initial start-up, it became apparent that the cleanroom pressures were not at all stable, as they were neither constant relative to each other nor relative to the warehouse space.
In response, the contractor attributed the instability to frequent unpredictable overhead door openings and to random, cycling economizer modes from the 20-odd general building HVAC systems in this multi-tenant space. In other words, the cleanroom pressures were unstable because the building pressures were not stable. The proposed solution was to provide and install a building pressure control system consisting of a differential static pressure transmitter—a through-the-wall exhaust fan with variable speed control. The system setpoint was +0.01” H2O relative to the outside (building exterior space).
The intent of the contractor’s solution was that as the building pressure increased due to the presumed unpredictable operation of its various HVAC systems, the exhaust fan speed would increase until the setpoint pressure was achieved. By doing so, the building pressure could be reliably controlled, and the cleanroom pressure fluctuations would disappear.
The following highlights the details of the differential pressure transmitter, DPT-1. The exterior static pressure tap was a 3/16-inch brass tube bent downwards at 90 degrees and exposed to the weather. When the wind would blow into the page, the velocity pressure of the air outside the building would obviously increase. This increase in velocity pressure at the open end of the brass tube would be interpreted as a decrease in static pressure at the open end of the brass tube since TP = VP + SP. That is also how a curveball works; the rotating stitches on the ball are like fan blades, increasing the velocity on the underside of the topspinning baseball.
This results in a decrease in the static pressure, and since the static pressure on the top is now higher than underneath, any object drops more than it would otherwise due to gravitational forces. In this case, the drop in the low static pressure input of the transducer is interpreted as an increase in the high-pressure terminal of the transmitter. The transmitter has no way of knowing if the high-pressure side increased or the low-pressure side decreased.
For the moment, the relationship between the cleanroom space and the warehouse space has not changed, and the pressures are relatively stable. However, the pressure transmitter is now sending a signal to the exhaust fan variable-frequency drive (VFD) to increase speed to counteract the increase in building pressure that it believes it is experiencing. Speeding up, the exhaust fan now lowers the interior building pressure. This results in the cleanroom pressure gauges reading between the cleanroom and the warehouse increasing until such time that the building pressure control system is satisfied and begins to slow back down.
When this happens, the building pressure now increases relative to the exterior, causing the difference between the cleanroom pressure and the building pressure to reduce. After a few seconds or minutes, another gust of wind appears, and the cycle continues. Meanwhile, the cleanroom pressure is being logged as unstable since it is constantly going up and down relative to the warehouse. The building exhaust fan solution has become an amplifier of the problem rather than an attenuator.
The eventual solution to the problem was conceptually simple but much more challenging to implement. The existing constant airflow control system, which was active on each of the two AHU’s supply air outlets, one of the two AHU’s return air ducts, one exhaust fan and one AHU face/bypass damper by variable frequency drives, were disconnected from automatic control, and the speed drives were manually adjusted to achieve the required airflow. Secondly, the building-pressure exhaust fan was decommissioned.
The next step was to select a representative critical room supplied by each of the two AHU systems (there were two zones with nine rooms total in reality) for active pressure control. The critical rooms’ pressure sensors were monitored relative to the ambient, adjacent warehouse space, and the signal was used to control the outside air quantity in a closed-loop arrangement via newly installed motorized outside air dampers. Afterwards, room-room relative pressures were manually balanced for each of the two zones.
Since final pressure balancing would have required multiple iterations of adjustment with progressively diminishing effects since the automatic outside air damper would be acting to maintain the critical room’s pressure during the process, a different approach was taken. Consequently, the OA damper control loop was initially offline, the OA set to a constant CFM value, and each zone’s rooms balanced with respect to each other with special attention paid to the critical room’s relationship to ambient. At the satisfactory completion of this process, the active pressure-control system was engaged. In actuality, either pressure-balancing strategy could have been employed.
In truth, the facility, consisting of nine rooms and two zones, is under a partially closed-loop pressure control system since the two critical rooms are under closed-loop active pressure control, and their respective adjacent spaces are in an open-loop relationship with the OA damper (like a toaster). A purely closed-loop system would have required nine room pressure sensors and nine actuating airflow valves (dampers). This option was not available for this facility. However, as Winston Churchill replied when asked about the rigors of old age – consider the alternative.
While this particular case was very interesting to the author, it was much less amusing to the client. As cleanrooms become less porous and room pressures come under finer scrutiny, design engineers will need to reconsider the wisdom of open-loop pressure control.
Looking to get NEBB Certified? Request your application today.