Case Study - Fume Hood Control

This article Reprinted from The Official Journal of ISPE

describes an PHARMACEUTICAL ENGINEERING® March/April 2003, Vol. 23 No. 2

innovative

Step-Variable Air Volume Fume Hood

design of a
fume hood
control system.
Control - A Case Study
by Sarla M. Patel, PE and Martin J. Wendel Jr., PE

T
Introduction cause of the equipment and energy required to
his article describes an innovative de- exhaust air through the fume hoods, and to
sign of a fume hood control system. This provide conditioned ventilation air to the labo-
design innovation combines (for a spe- ratory spaces.
cific chemical development laboratory One of the many important laboratory de-
facility) the best features of traditional con- sign issues associated with fume hoods is the
stant volume and variable volume fume hood selection of a fume hood control strategy, which
controls, while avoiding some of their perceived then leads to the control strategy for overall
disadvantages. laboratory airflow and pressurization. Con-
stant volume fume hood control has been used
Design Challenge - in many laboratories for years, and is generally
Laboratory Fume Hood Control considered to be a solid, reliable approach to
Fume hoods are critical elements in the design safeguard workers. For laboratories with large
and operation of a pharmaceutical research numbers of fume hoods, constant volume fume
and development laboratory. Fume hoods are hood control results in enormous energy use
essential in many applications to protect work- and operating costs since the same amount of
ers from exposure to chemical fumes or other air must be conditioned and exhausted through
hazardous substances. At the same time, fume the fume hoods whether they are in use or not
hoods have a significant impact on the labora- (that is, whether the fume hood sashes are open
tory construction costs and operating costs be- or closed). More recently, variable volume fume
hood controls have
Figure 1. Bench top
been developed and
fume hood - shown with
sashes closed. installed to reduce
energy consumption
and operating costs in
fume hood-intensive
laboratories by reduc-
ing air flow through a
fume hood when the
sash is closed. Many
of these systems,
which typically main-
tain a constant fume
hood face velocity to
capture fumes and
protect laboratory
workers, also are gen-
erally considered to be
solid, reliable systems
with good track
records.

©Copyright ISPE 2003 MARCH/APRIL 2003 PHARMACEUTICAL ENGINEERING 1

Case Study - Fume Hood Control
The technical issues associated with fume hood control hood sashes will be open at any one time. Since the high fume
strategy selection can be quantified and analyzed in terms of hood density in this laboratory meant that the room air change
performance, installed cost, operating, and maintenance costs. rate was exhaust-driven, the diversity also permitted reduc-
However, it is essential that this system selection (and most tions in supply air conditioning system and chilled water
others) includes consideration of owner acceptance, including system sizing. The fume hood operating diversity was based on
the scientific research community that will use the labs and the agreement by the scientific researchers to institute a
the building engineers who will maintain the systems. The program of fume hood sash management so that the maximum
building design must be responsive to the needs of the people combination of hoods with sashes open at the same time (on a
who will work in the building. Sometimes, as a result of the per-lab basis) would be two bench top hoods and two walk-in
consideration of both human and technical requirements, the hoods, or four bench top hoods and one walk-in hood.
building design professional has to think outside the con-
straints of pre-packaged solutions to achieve design excel- Design Criteria
lence. This article discusses one example of this type of The solution to the design challenge had to meet the following
innovation, and the benefits of the resulting system design. criteria:

The Case Study 1. The solution must maintain safe conditions for scientific
The building design project was a 75,000 gross square foot researchers and other laboratory workers.
(6,970 gross square meters) chemical development facility.
The scientific work planned for this facility required fume 2. The solution must satisfy user demands for simplicity of
hood-intensive laboratory spaces. A typical 735 square foot instrumentation, controls, and operations.
(68 square meters) laboratory module was planned to include
a total of six fume hoods, four bench top hoods, and two walk- 3. The solution must permit reduction in fume hood exhaust
in hoods. airflow to allow economic HVAC system design based on
The operating profile of the fume hoods was identified by operating diversity.
the hood users as generally being one or the other of two
scenarios: sashes either 50% open or fully closed. The hood 4. The solution must permit further reduction in fume hood
sashes would be opened 50% either when experiments are exhaust airflow during operation to lower energy con-
“setup” within a hood, or certain operations are performed sumption and costs without a substantial maintenance
that require the user to frequently be in-and-out of the hoods. cost penalty for ongoing component and system calibra-
At all other times, the sashes would be kept closed. Situations tion and adjustments.
where the sashes would be partially open at a point some-
where between 50% and fully closed were not considered a Traditional Design Alternatives -
“real world” condition that would occur in practice. CAV and VAV Control
The building design professionals proposed to design Heat- The two design alternatives traditionally considered for this
ing, Ventilating and Air-Conditioning (HVAC) systems ac- type of laboratory facility are Constant Air Volume (CAV)
counting for diversity in fume hood use to achieve construction control, which utilizes CAV boxes controlled to satisfy a
and operation savings. That is, if fume hood exhaust can be constant airflow setpoint, and Variable Air Volume (VAV)
reduced when the sashes are closed, the exhaust systems can control, which utilizes VAV boxes configured to provide
be sized for a diversified load since only a fraction of all fume variable system airflow. Since the ultimate design solution
included a mix of the concepts involved in these two tradi-
tional designs, their application will be briefly reviewed.

Constant Volume Fume Hood (Figure 2)

Constant volume bypass hoods (CAV) typically feature hood
sashes that operate in tandem with a “bypass” that is reverse
acting with the position of the sashes. This arrangement
provides a constant exhaust flow, measured in cubic feet per
minute (cfm), through the fume hood regardless of the sash
positions - Figure 2. The exhaust airflow through the hood is
maintained constant through the use of a CAV box installed
in the ductwork serving the hood that is calibrated and
controlled to always provide a constant air volume CAV
exhaust flow setpoint. This CAV box has an airflow measur-
ing device and associated controls that modulates an “air
valve” the same way a VAV box is configured. The only
difference is that with a CAV box, the “control system”
Figure 2. Constant volume bench top fume hood. modulates the CAV box to maintain a programmed constant

2 PHARMACEUTICAL ENGINEERING MARCH/APRIL 2003 ©Copyright ISPE 2003

 Case Study - Fume Hood Control

“Sometimes, as a result of the consideration of both human and technical

requirements, the building design professional has to think outside the constraints of
pre-packaged solutions to achieve design excellence.”

flow, regardless of “upstream” system dynamics in airflow 3. manually via a relay wire to the light switch in the space
and pressures. If we look at a typical CAV bench top fume where the hood is installed to index the hoods to the
hood, when the hood sashes are fully open, the bypass is fully occupied mode when the lights are turned on and the
closed, and the air flows through the sash opening (and unoccupied mode when the lights are turned off.
airfoil). When the sashes are fully closed, the same quantity
of air flows through the bypass (and airfoil). The CAV boxes 4. automatically via motion sensor technology detecting the
are applied to the system to draw a constant exhaust airflow presence of a researcher
rate from the hood regardless of whether the exhaust air
flows through the sash opening or the bypass. 5. automatically via monitoring the position of the sashes to
CAV hoods are high energy users because there is no detect sash closure - unoccupied when hood sash is fully
reduction in exhaust air out of the hoods when the sashes are closed and occupied when hood sash is not fully closed
partially or fully closed. No reduction in exhaust air also
means no reduction in conditioned supply (makeup) air Variable Air Volume (VAV) Hoods
delivered to the lab to balance the exhaust air from the hood. In VAV fume hoods, the airflow through the hood varies
One energy savings alternative typically applied to CAV depending on hood sash position or hood differential pres-
hoods is to index (or “set back”) the hoods to a lower exhaust sure. Figure 3 depicts airflow with hood sashes open and
flow when the laboratories are unoccupied. During this unoc- closed respectively. In VAV hoods, the exhaust airflow through
cupied mode, the exhaust air through the hood is reduced to the hood is typically varied to maintain a constant face
some percentage of full flow (60% for example), which also velocity through the hood sash opening through the use of a
results in a direct reduction of supply air to the space where VAV box installed in the ductwork serving the hood. The
the hood is installed. This flow set back approach has typi- supply and exhaust airflow is controlled by sensing the hood
cally been applied on a lab-by-lab basis and sometimes sash position or by sensing airflow via an anemometer lo-
requires additional construction expense to provide an occu- cated in the hood side wall.
pied/unoccupied indication outside of the laboratory. Hood sash position control utilizes a potentiometer or
One or a combination of the following methods may control similar position-sensing device attached to the sash through
the occupied/unoccupied mode changeover: a cable. When the hood sash position changes, the sensing
element responds to the change in sash position. The associ-
1. globally for all laboratories via a time schedule ated controller feeds back to the Building Management Sys-
tem (BMS) to determine the required air quantity and control
2. manually via a hand switch provided for each hood or of the exhaust air regulator to achieve a position (and corre-
group of hoods sponding exhaust airflow) that corresponds to the new sash
position.
Hood side wall anemometer control utilizes an airflow
sensor called a thermal anemometer to infer hood face veloc-
ity. The sensor measures air velocity passing through a hole
in a side wall of the hood, which has been shown experimen-
tally to represent an average face velocity. When the hood
sash is opened, the sensor measures the reduction in face
velocity and sends an appropriate signal to the exhaust air
regulator, via the BMS, to increase the airflow until the face
velocity setpoint is reached. When the hood sash is lowered,
the sensor measures a velocity increase and sends a signal to
decrease airflow.

Comparing Traditional CAV

and VAV Fume Hood Control
Comparing traditional CAV and VAV fume hood control, one
may observe that CAV control is free of complex instrumen-
tation and hence inherently more reliable. The hood operat-
Figure 3. Variable volume bench top fume hood.

©Copyright ISPE 2003 MARCH/APRIL 2003 PHARMACEUTICAL ENGINEERING 3

Case Study - Fume Hood Control
Hood Type Sashes 50% Open All Sashes Fully Closed module was a “hood driven” space. This means that even
when all of the hoods were closed, the minimum combined
Bench Top 1140 CFM (538 L/S) 475 CFM (224 L/S) airflow setting for all hoods with all sashes closed was a value
Walk-In 2220 CFM (1048 L/S) 890 CFM (420 L/S) that required “make-up” supply air flow from the air handling
system that was beyond that required to offset the heat gains
Table A. Fume hood exhaust volumes.
in the laboratories.
ing integrity is maintained because the required amount of The hood setback exhaust airflow quantity was found,
air is available all the time. The primary disadvantages of through factory smoke testing, to provide good mixing condi-
CAV control are the energy cost of exhausting unused condi- tions in the hood with air flowing through the bypass in the
tioned air when the sashes are closed, and high air handling low flow condition.
unit and ductwork costs since no diversity can be taken with
CAV. Cost Comparison
Traditional VAV control provides more economical opera- The first cost for a constant volume hood control installation
tion when the sashes are less than fully open as it takes only is significantly less than that of a variable air volume hood
the amount of conditioned air required to maintain a mini- control installation for the same size hood. Table B is a
mum face velocity with minimal bypass. Constant face veloc- summary of the major hardware elements along with “aver-
ity is maintained and continuous monitoring and alarming of age” associated costs that comprise a typical CV and VAV
face velocity is provided. However, there is an additional cost control system installation for a lab module featuring two
of instrumentation for sensors and controllers. Furthermore, fume hoods. The HVAC system serving the model profiled
VAV control sometimes requires a separate proprietary fume here has a dedicated “exhaust box” for each hood, and a single
hood control system, independent of (but usually interfaced supply air box that provides makeup air for the hoods and
to) the BMS. This increases installed cost and operating/ conditioning for the space.
maintenance costs for the system.
Building Management System
The Design Solution and Instrumentation Requirements
The design solution, referred to as Step-VAV control, is a The microprocessor-based Direct Digital Control (DDC) Build-
combination of the best features of traditional CAV and VAV ing Management System (BMS) monitors and controls the
controls, while avoiding some of their perceived disadvan- lab/hood air flow and space temperature. This system is an
tages. extension of the BMS used for total building HVAC system
The design solution utilized the fume hood types preferred control and monitoring. This Step-VAV control design does
by the researchers - Figure 1. The hoods were bypass CAV not require a proprietary lab airflow control system.
with taller than standard height hood sashes. Each hood was Instrumentation requirements for laboratories using Step-
provided with four horizontal sliding sashes in a double track VAV control are very simple compared to labs using tradi-
arrangement so that no more than 50% of the hood sash area tional VAV control. Each exhaust box utilizes the box supplier’s
could ever be open at any one time. Whenever a hood sash is standard pneumatically operated two-position control prod-
opened, its bypass door is closed in the same proportion. Each uct. When pneumatic control signal pressure is 0 psig (101
sash door is furnished with a door switch. The door switch kPa), the box is at maximum exhaust CFM, and when
contact closes when the hood sashes are fully closed. The pneumatic control signal pressure is 13 psig (224 kPa), the
contact opens when any sash is moved from the fully closed box is at minimum exhaust CFM. The pneumatic signal is
position. As noted, there are four doors at each hood. Since the provided by the BMS, 13 psig (224 kPa) when all sashes at the
control strategy is based on detecting when all the sashes are
fully closed, the position of individual sashes is not important Constant Volume Hood Control Installation
and the fume hood manufacturer wired all switches to pro-
Terminal exhaust box controller $2,000
vide one common contact closure per fume hood. Each hood is
provided with a low flow alarm light, supplied by the hood Terminal supply box controller $2,000
manufacturer. The low flow alarm and auxiliary alarm con- Hood alarm $1,000
tacts are activated whenever hood face velocity drops below Unoccupied/occupied override switch $500
the set flow. Hood airflow cfm (cubic feet per minute) require-
Total $5,500
ments are indicated in Table A.
The laboratory exhaust system is comprised of constant
Variable Air Volume Hood Control Installation
volume two-position air terminal boxes, one per hood. The
exhaust is ducted from each hood directly to its dedicated Fume hood controller with sash sensor, hood alarm panel $8,000
and exhaust box control
exhaust box. The laboratory supply system is comprised of
two VAV supply boxes that serve each lab module. There is a Lab control panel with flow tracking, supply box and $7,000
room temperature control
reheat coil down stream of each VAV box. Air is delivered to
the room through diffusers. Total $15,000
Lab general exhaust was not required as the laboratory Table B. CAV/VAV hood control installation cost comparison.

4 PHARMACEUTICAL ENGINEERING MARCH/APRIL 2003 ©Copyright ISPE 2003

 Case Study - Fume Hood Control
hood are fully closed, and 0 psig (101 kPa) when one or more alty to provide an individual RCP per lab (compared to
sashes is not fully closed. controlling several labs from one RCP) was not prohibitive.
Control response time of the instrumentation used to Each supply box included an airflow probe (furnished by the
index the setback was found to be fast enough to maintain box suppler), a differential pressure-type airflow transmit-
hood containment while switching the exhaust boxes be- ter, a damper (furnished by the box supplier), and an electric
tween modes. damper actuator.
Supply boxes are controlled by the BMS supplier’s stan- The reheat coil valve actuator is electrically operated.
dard product called a “Remote Control Panel (RCP).” This Space temperatures are measured by wall mounted tempera-
panel is a stand-alone controller and communicates on a ture sensor/transmitters. There are also two duct mounted
“peer-to-peer” basis with other RCPs over a dedicated com- temperature sensors/transmitters, one in each supply duct.
munication network. For reliability, one RCP was provided The reheat coil valves are controlled using space temperature
per lab. In the event of failure of a RCP, only one lab would be sensors. The duct mounted temperature sensors are used for
affected. Because the RCPs have a modular design and were anticipatory BTU control of the reheat control coil valves –
sized to provide required input/output points, the cost pen- Figure 4.

Figure 4. Typical Step - VAV Lab Module Controls.

©Copyright ISPE 2003 MARCH/APRIL 2003 PHARMACEUTICAL ENGINEERING 5

Case Study - Fume Hood Control
Number of Walk-In Hoods with Number of Bench Top Hoods with Laboratory Air Flow Volume
Sashes 50% Open <airflow> Sashes 50% Open <airflow> for this Configuration
2 <total 4,440 cfm/2.096 L/s> 2 <total 3,230 cfm/1,525 L/s> 7,670 cfm (3,620 L/s)
2 <total 4,440 cfm/2.096 L/s> 1 <total 2,565 cfm/1,211 L/s> 7,005 cfm (3,306 L/s)
2 <total 4,440 cfm/2.096 L/s> 0 <total 1,900 cfm/897 L/s> 6,340 cfm (2,992 L/s)
1 <total 3,110 cfm/1468 L/s> 4 <total 4,560 cfm/2,152 L/s> 7,670 cfm (3,620 L/s)
1 <total 3,110 cfm/1468 L/s> 3 <total 3,895 cfm/1,838 L/s> 7,005 cfm (3,306 L/s)
1 <total 3,110 cfm/1468 L/s> 2 <total 3,230 cfm/1,525 L/s> 6,340 cfm (2,992 L/s)
1 <total 3,110 cfm/1468 L/s> 1 <total 2,565 cfm/1,211 L/s> 5,675 cfm (2,679 L/s)
1 <total 3,110 cfm/1468 L/s> 0 <total 1,900 cfm/897 L/s> 5,010 cfm (2,365 L/s)
0 <total 1,780 cfm/840 L/s> 4 <total 4,560 cfm/2,152 L/s> 6,340 cfm (2,992 L/s)
0 <total 1,780 cfm/840 L/s> 3 <total 3,895 cfm/1,838 L/s> 5,675 cfm (2,679 L/s)
0 <total 1,780 cfm/840 L/s> 2 <total 3,230 cfm/1,525 L/s> 5,010 cfm (2,365 L/s)
0 <total 1,780 cfm/840 L/s> 1 <total 2,565 cfm/1,211 L/s> 4,345 cfm (2,052 L/s)
0 <total 1,780 cfm/840 L/s> 0 <total 1,900 cfm/897 L/s> 3,680 cfm (1,737 L/s)
Table C. Hood sash/lab air flow combinations.

Control Sequence of Operations Operating Reduction in Airflow

1. The BMS monitors the hood sash position for each fume The system is designed to accommodate a maximum of two
hood via sash end switch contact. bench top hoods and two walk-in hoods, or four bench top
hoods and one walk-in hood in simultaneous operation with
2. The BMS changes each exhaust box CFM setpoint based sashes 50% open. Since the hood sashes are typically opened
on its sash position. for experimental setup and closed otherwise, the Step -VAV
design permits significant flow reduction within each lab.
3. The BMS calculates total exhaust CFM by summing air The required lab air flows are the sum of the flows for hoods
exhausted from each hood with sash open and each hood with sashes 50% open plus flows for the hoods with all sashes
with sash closed. Using the result, the DDC System closed. For example, with one walk-in hood open, the lab air
calculates required supply air CFM (supply CFM = ex- flow for two walk-in hoods is 1 x 2,220 cfm (1048 L/s) + 1 x 890
haust CFM – constant transfer CFM). As there are two cfm (420 L/s) = 3,110 cfm (1468 L/s). The different hood
supply boxes serving each lab module, each box will operating conditions and corresponding airflow volumes are
provide half of the total required CFM. The BMS modu- indicated in Table C.
lates the supply box dampers to maintain the supply CFM As can be seen by inspection of the values in Table C, Step
setpoint. VAV control for the typical labs in this case study provides
stepped modulation of air flow from a maximum of 7,670 cfm
4. The BMS limits supply CFM to the maximum design limit (3,620 L/s) to a minimum of 3,680 cfm (1,737 L/s) in 665-cfm
- to allow a maximum of two bench top hoods and two walk- (314-L/s) increments. In this application, Step VAV control
in hoods open at the same time, or four bench top hoods and provides the same operating cost benefits as a traditional
one walk-in hood. If a researcher opens sashes in more VAV control approach. Since the researchers typically open
than the design allowance of fume hoods, the BMS acti- their hoods for experimental setup and keep them closed at
vates a strobe light in the lab indicating “TOO MANY other times, the lab airflow demand is stepped by its nature
HOODS OPEN – CLOSE HOOD SASHES.” and any additional airflow reductions provided by traditional
VAV control would be only transitory.
5. If hood sashes are not closed immediately upon activation
of the strobe light, exhaust air from each hood reduces, System Benefits
resulting in a further drop in face velocity. When the face The design solution implemented for this project provides
velocity drops to its preset low limit, a “LOW FACE significant benefits to the building owner, including the
VELOCITY” alarm light at each hood (provided by hood following:
manufacturer) is activated.
1. This design does not require the complex instrumentation
6. The BMS maintains space temperature setpoint by modu- and control of conventional VAV control systems, such as
lating reheat coil control valves from the signal of one a VAV exhaust box controlled from hood sash position and
space temperature sensor and two duct mounted tempera- tracking supply with each change in each hood sash
ture sensors, one in each discharge duct in cascade mode. position. Monitoring individual hood sash positions for the

6 PHARMACEUTICAL ENGINEERING MARCH/APRIL 2003 ©Copyright ISPE 2003

 Case Study - Fume Hood Control
horizontal sash arrangement with four doors is not re- hood is controlled in two steps; maximum flow when any sash
quired. is open, and minimum flow when all hood sashes are closed.
In the present application of six hoods, there are 13 stepped
2. The first cost of Step VAV is significantly less than the first flow conditions, providing a stepped approximation of VAV
cost of traditional VAV for the same fume hood applica- control. Supply flow is controlled by the BMS to maintain
tion. transfer airflow directionality. The installed cost of this
system is lower than traditional VAV and the resulting
3. There should be no differential cost to the Owner in the design is very flexible.
programming of Step VAV versus traditional VAV as the
BMS system would have to be programmed regardless of About the Authors
the HVAC operating scheme. Sarla M. Patel, PE is a Senior Project Engi-
neer at Kling. Patel has more than 35 years
4. This design is more robust than traditional VAV control of experience in instrumentation and con-
because it is more tolerant of variations in component trol, electrical engineering, reliability, main-
(such as exhaust damper) response times. Rather than tainability, and computer microprocessor ap-
trying to “chase” traveling sashes, this design provides full plication. In her present position, she is re-
exhaust flow as soon as any sash is moved from the fully sponsible for project engineering, design and
closed position, and maintains full exhaust flow until all development, and for equipment selection.
sashes are fully closed. Patel’s experience includes systems for automatic tempera-
ture control, building automation, energy monitoring and
5. Airflow control is provided by the time proven technology control, security, fire alarm and smoke detection control
of the BMS (DDC System), which is simple, cost effective, systems, motor control centers, sound and communications,
and reliable, without introducing additional lighting, data communications, telephone electronics, pro-
manufacturer’s systems (that could add complexity and cess control, and local area networks. She has provided the
cost to building maintenance) into the facility. engineering services for Glaxo Inc.’s Pharmaceutical Head-
quarters Campus in Research Triangle Park, NC, Wyeth
6. It is anticipated that overall maintenance costs (typically Research Inc.’s Great Valley Corporate Center in Frazer, PA.,
averaging approximately $100/per sensor) will be less for and multiple research laboratory renovations for Merck and
Step VAV as there are fewer BMS points associated with Co. Patel is a member of the Instrumentation Society of
this scheme over traditional VAV applications. America and the Institute of Industrial Engineers. She holds
a bachelor’s degree in electrical engineering from the Univer-
7. In this application, where the hood sashes will generally sity of Michigan and a master’s in industrial engineering
be opened only while setting up experiments within a hood from the University of Pittsburgh. She can be contacted at:
and sashes will be kept closed at all other times, step VAV spatel@[Link].
control is more appropriate to the owner’s intended opera-
tion compared to conventional VAV. Capital cost savings Martin J. Wendel, Jr., PE is an Engineer-
and operating/energy cost savings are achieved through ing Design Principal at Kling. Wendel is
the simple addition of a single point, “go/no-go” measure- responsible for engineering leadership and
ment at each fume hood. Put another way, if a traditional oversight of all mechanical services disci-
VAV system were designed into the case study facility, it plines, process plumbing, fire protection, and
would be used in a manner that profiles the way a “Step- instrumentation and controls. Wendel has
VAV” system would be used, so the operating costs would expertise on a wide variety of office, research
be similar. The key is, why pay the high first cost of and development, and computer facilities,
traditional VAV if you are not going to use the system that having designed mechanical systems for high-technology
way? The true savings are lower first cost with “Step-VAV” laboratories, pharmaceutical development facilities, US Fed-
control over traditional VAV control. eral Government buildings, and manufacturing facilities. He
has provided professional services for Merck and Co. in
Summary Wilson, NC, Glaxo Inc.’s Pharmaceutical Headquarters Cam-
The implemented hood control design is Step-VAV providing pus in Research Triangle Park, NC, and Wyeth Facilities in
only the required amount of air to each hood. This minimizes Pearl River, NY. Wendel is a member of the American Society
the capital cost (through diversity in equipment sizing) and of Heating, Refrigeration and Air Conditioning Engineers,
operating costs. The control system is simple, reliable, and and ISPE. He holds a bachelor’s degree in mechanical engi-
easy to maintain. It utilizes the researchers’ preferred con- neering from Drexel University. He can be contacted at:
stant volume bypass hoods. The only automatic sensing mwendel@[Link].
requires hood sash end switch position monitoring by the Kling, 2301 Chestnut St., Philadelphia, PA 19103, (215)
BMS, using simple proven technology. Exhaust flow for each 569-2900.

©Copyright ISPE 2003 MARCH/APRIL 2003 PHARMACEUTICAL ENGINEERING 7