Waste Anesthetic Gas Disposal

(WAGD) Systems

Mark Allen

TM

Continuing Education Publication

Notes

Notes on Using this Pamphlet:

This pamphlet is presented as a service to systems designers and operators selecting, designing and working with
piped WAGD systems in medical facilities. The pamphlet seeks to improve understanding of what these systems
are intended to acheive, how they interface with medical devices, what safety risks they pose, and what options
are available under the NFPA 99 “Health Care Facilities” standard.

The 2002 version of the NFPA 99 is used for all references except where indicated. Users are cautioned that this
pamphlet is intended to be used in conjuction with the standard, which should be obtained from:

National Fire Protection Association

1 Batterymarch Park
Quincy, MA 02269-9101
Phone 1-800-344-3555
Internet www.NFPA.org

Preliminary Edition 12 May 2004

Comments on this booklet or on any aspect of medical gases are welcome and encouraged.
Please send to mallen@beaconmedaes.com

This Pamphlet in both print and electronic versions is Copyright 2004 BeaconMedæs and Mark Allen. All Rights are
Reserved, and no reproduction may be made of the whole or any part without permission in writing. Distribution
of the Electronic version is permitted only where the whole is transmitted without alteration, including this notice.

WAGD Guide 3
 Contents

Table of Contents Some terms used in this booklet

Some Terms …………………………………… 4 AGSS : “Anaesthetic Gas Scavenging System”. The

International Standards equivalent to WAGD (Waste
Introduction ………………………………… 5 Anesthetic Gas Disposal) as used in the U.S.
The issues and why WAGD is suddenly a hot topic.
Dedicated Implementation : A system with a producer
exclusively for the WAGD system. The opposite of a
An Anesthesia and WAGD Primer ……… 7 Dual Use implementation.
What a WAGD system does and how it interfaces
with the Anesthesia machine. Dual Use : A combined system of medical vacuum and
WAGD.
The view from the NFPA 99 …………….… 10
Options for WAGD under the standard. EN System : A system which complies in it’s essentials
with the European AGSS standard EN 737-2:1998
The Ideal WAGD System …………………….… 10
Critical elements to consider in selecting an High vacuum : As used herein, a vacuum above 5
inHgV (127 mm HgV). Typical high vacuum systems
implementation.
are run at medical vacuum levels 15-29 inHgV (380
- 760 mmHg)
Existing System Upgrades …………………… 11
How to deal with a system already installed. Intrinsically Safe : A system which is designed so that
no failure or combination of failures can cause the
Overview of WAGD Implementations …… 14 system to endanger the treatment or the patient.
All the WAGD implementations explained.
Low vacuum : As used herein, a vacuum below 5
Selecting an Implementation ……………… 16 inHgV (127 mmHgV)
All the WAGD implementations compared.

General requirements ………………… 17

What NFPA requires in any WAGD Implementation.

Common Questions …………………………… 18

Sources for additional material …………… 19

Where to get additional information.

4 WAGD Guide
Introduction

Introduction enough, that few if any serious problems of this type have
been reported.
Why a booklet on such an obscure topic? Unique among
medical gas and vacuum systems, failure of a Waste The average engineer designing medical gases is largely
Anesthetic Gas Disposal (WAGD) system does not usually untutored in the function, usage, internal complexities
pose a life safety hazard to patients. WAGD is reasonably and underlying hazards of a WAGD implementation.
forgiving, and even a marginal implementation often can The result has been that North American medical gas
be induced to operate adequately. WAGD deals with an designers have largely come to share this common set of
occupational hazard - not a problem usually considered assumptions:
part of medical gas work. Indeed the whole topic and
the systems themselves might long ago have vanished 1. Dumping the WAGD into the medical vacuum source
but for NIOSH/OSHA mandates for testing trace gases is cheap. This is assumed to be true in equipment terms
in occupational workplaces, including operating rooms. and also in the designer’s own time - no line sizing or
Only because WAGD is the easiest way to control pump sizing needs to be done other than what must be
these trace gases does WAGD remain part of the O.R. done for the vacuum anyway.
landscape.
2. In a Dual use implementation, any oxygen or other
The whole subject remains controversial. There are gases will arrive at the pump sufficiently diluted to render
many who would contend that waste gas presents no them harmless.
proven health hazard and that no adequate study exists
to justify the continuing testing, much less the installation 3. Any other problem will be handled by someone else.
of WAGD in every O.R. However, given even casual
consideration, commonsense must prevail: These assumptions have lead to the most common
implementation of WAGD in North America today - a
“It only makes sense that if a potent anesthetic WAGD terminal at each anesthetizing location piped into
gas causes individuals to slump into a semi the medical vacuum line and thence into the medical
comatose sleep before they can count from ten vacuum pump.
to one backwards, incremental amounts of the
same anesthesia … would cause similarly acute In no area of medical gases are international practice and
effects in nearby surgical personnel.” North American practice more widely divergent. Unlike
the loose approach taken in North America, Europeans
“Reducing Exposure to Nitrous Oxide have put much effort toward perfecting these systems.
Thom Wellington, The result of this work is found in the standard (EN 737-
article, FacilityCare Magazine, 2) used in Europe, which includes some very interesting
August 1998 advances:

Achieving this very commonsense goal creates hazards 1. EN systems are designed to be intrinsically safe. That
of it’s own. Two in particular have exercised us in the means that the system protects the patient, no matter
past: The first was much discussed when WAGD systems what the anesthesiologist does or doesn’t do. EN systems
(usually termed “Evacuation” or “Evac” in those days) first cannot expose the breathing circuit to the intense vacuum
came into widespread use. There was much discussion of the medical vacuum system.
about the consequences of drawing oxygen and nitrous
oxide into oil lubricated pumps. Although the concern 2. EN AGSS is always implemented with a dedicated
was theoretically valid, such implementations became less producer and a dedicated piping network. They do not
and less unusual and the fires or explosions predicted did use any components which are incompatible with the
not materialize. It is now true that some major engineering waste gases or with oxygen.
firms take no notice of this hazard at all and routinely
specify Dual use implementations through lubricated 3. Piped systems are designed around low cost, low
machines. horsepower regenerative or vortex blowers instead of
relatively expensive pumps. This makes the systems
The second hazard is the risk of applying full line relatively inexpensive (no small consideration in today’s
vacuum to the anesthesia breathing circuit. Here too, environment.)
the theoretical risk appears to not have proven to be the
real problem originally feared. The interface valves on It appears that these EN assumptions about WAGD are no
the anesthesia machine have proven reliable enough, or longer merely of curious interest. The major anesthesia
the intervention of the anesthesiologist has been swift machine manufacturers operate globally and these

WAGD Guide 5
 What is Happening

manufacturers In older machine designs, this drive gas is vented into

have begun to the room. This gas may in many cases be pure oxygen,
bring anesthesia and of course may be flowing the entire length of the
systems originally case, greatly increasing the total volume of oxygen
designed using entering the WAGD system. The potential for exposure
these European of the pump to elevated oxygen levels is therefore
assumptions into greatly increased.
North America.
Two unexpected The claim is made that Dual use systems are safe
consequences are because they enjoy “dilution’ - that is, the waste gas
Fig. 1 being reported: is diluted by the air entering the standard vacuum
A pump used in Dual Use inlets. NFPA even lends credence to this argument
service which appears to 1. Fires in vacuum in 5.1.3.7.1.2. The argument has been impossible to
have suffered an oxygen fire. pumps. We had no prove or disprove, because heretofore no evidence has
verifiable reports of existed either way, except an inconclusive absence of
fires in WAGD pumps for many years. But beginning in reports of fires in such systems.
the last half of 2002 and continuing through today, there
have been increasingly frequent reports of pump fires. There has recently been undertaken a study wherein
Such reports are hard to corroborate, and many reports the oxygen levels at the exhaust of a pump in Dual use
we have not been able to document, but we have been service were actually measured. The study conclusively
able to verify some others. (see Fig. 1). demonstrated that oxygen levels did in fact rise to
dangerous levels (35%) when these new anesthesia
The reports range from flashes at the exhaust through machines were in service. While the effect of these new
to complete explosive destruction of pumps. In all the anesthesia systems in any given WAGD system will vary,
cases corroborated so far, the pumps were in dual use it is clear that they can and do elevate the level of oxygen
WAGD and medical vacuum service, and all were oil passing through these pumps. (Scott, et al, 2004) We
lubricated. therefore now can conclude that while dilution may be
present in any given system, it is not a reliable substitute
2. Vacuum pumps running excessively. In numerous for a proper WAGD implementation.
cases both with vacuum pumps in Dual use service and
in dedicated WAGD service, pumps have been seen to How dare they!
be running much harder than expected or than that same
pump had run historically. In extreme cases, this has It is very important to caution the reader that there is
resulted in pump failure. no proof of a direct cause-effect between these new
anesthesia machines and these adverse effects, and that
We have been able to determine a possible set of causes all evidence is at this writing circumstantial. The reader
for these reports. These bear directly on the assumptions should use these reports and the information contained
under which WAGD is designed and installed: in this paper as a guide to making improvements in their
WAGD systems in order to increase the absolute safety
1. There are now on the US market some anesthesia and reliability of their systems. It would be entirely
systems with interface valves which require larger inappropriate and irresponsible to use this information
inflows than previous machines. to influence decisions regarding medical practice or the
selection of any given anesthesia system, which is an
The ‘traditional’ interface valve would draw in the 6-9 entirely medical decision.
liters per minute range when in operation. Under EN
standards, the inlets are required to draw 50 lpm (1.8 While at first blush it may seem irresponsible for the
scfm). However, the standard sizing assumptions used anesthesia machine manufacturers to make these
in North America for WAGD (for instance in the NFPA changes without notice, it’s important to understand
99 vacuum sizing methodology) assume only 1 scfm that what they have done is entirely consistent with the
(28.3 lpm) per anesthetizing location. rules for WAGD as NFPA has them today.

2. Some anesthesia systems are now venting their NFPA has required WAGD systems be compatible with
ventilator drive gas (the gas used to power the ventilator oxygen for many years. Although it has been watered
itself and not breathed by the patient) into the waste gas down somewhat over the years, NFPA 99 still carries
stream. a warning about the mixing of WAGD and medical
vacuum (ref: NFPA 99, 2002 5.1.3.7.1.2(2), A-5.1.3.7

6 WAGD Guide
The Anesthesia Machine

and also see NFPA Health Care Facilities Handbook, inHgV, but for the medical professional, it would be 760
2002 edition, commentary on 5.1.3.7). NFPA does not mmHgV. That same perfect vacuum could lift a column
stipulate an inflow, but does require that the pump used of water 405 inches. An inch of mercury is therefore
in such a dual use system “be adequate to handle the roughly 25 mm of mercury or 13.5 inches of water.
volume” (ref: NFPA 99, 2002 5.1.3.7.1.2(3)). The new
machines clearly will work very well and entirely safely Figure 1 is a simplified diagram of an anesthesia system.
on any system which complies with these rules. The breathing circuit is a closed loop with the patient
connected at the end. Gas from the piped medical gas
Venting the anesthesia machine ventilator drive gas systems enter the machine through the flowmeters and is
into the waste gas stream is done for a significant blended with an anesthetic agent in the Anesthetic Agent
safety reason. It does sometimes occur that there is Vaporizer. It is the proportioning of these gases and the
leakage from the breathing circuit past the ventilator anesthetic drug under the control of the anesthesiologist
bellows. In that case, the ventilator drive gas discharge which induces and maintains the desired level of
can contribute to anesthetic gas levels in the O.R. By anesthesia.
discharging the drive gas into the waste gas stream, this
possible source of contaminant is eliminated. The gas mixture passes into the breathing circuit through
the fresh gas inlet. When the patient breathes in, the
Regrettably, operators of WAGD systems are not gas is drawn through the inhalation valve, through the
asking (indeed don’t know to ask) questions about circuit tubing and inhaled.
the anesthesia systems that are being purchased. As
these newer machines come on line, the first anyone As the patient breathes out, the gases pass through the
may know of the change is when the facility begins other arm of the circuit, through the exhalation valve
experiencing pump problems. The problems will and into the rebreathing bag.
worsen as these new machines replace older anesthesia
systems still in use. When the patient inhales again, the gases are drawn
from the rebreathing bag, through the CO2 absorber to
A natural question might be “If I just don’t use this scrub out excess CO2 and the gas then passes back up
new machine, can I avoid any WAGD complications?” through the inhalation valve, mixing with fresh gas on
Obviously, the answer may be yes, for a while. But it’s way back to the patient.
the simple home truth is that mechanical systems in a
hospital are meant to facilitate the best medical practice, Since the circuit is closed, it is obvious that gas cannot
not to dictate it. be added indefinitely but must eventually be vented
somehow. This is the function of the Adjustable Pressure
It is thus essential that we change our languid and Limiting Valve (APL valve), which is set by the anesthetist
comfortable assumptions about WAGD, both in new to maintain a certain pressure in the breathing circuit
designs and even in existing installations. and to relieve any pressure in excess of that setting. The
gas which is vented through that valve is the waste gas
It is the purpose of this pamphlet to offer some guidance which our WAGD system will be expected to remove.
on this very pressing issue. Before we begin, a short It contains whatever mixture is present in the breathing
primer on anesthesia systems, WAGD and how the circuit, including halogenated anesthetic drugs, nitrous
systems fit together may be helpful. oxide, air, oxygen, water vapor, and carbon dioxide.

The Anesthesia Machine (ref. Figure 2) It is important to note that there is a very slight (inches
of water at most) positive pressure at the APL valve.
We will not even attempt to address all the complexities The pressure at the APL valve is crucial to the proper
of a modern anesthesia system here, but rather will look functioning of the whole anesthesia system. If a
at a system reduced to those basic elements which bear vacuum were present at the APL valve, it could suck
on how waste gases come to be and what they look like the anesthestic gases out of the breathing circuit. (In
to the WAGD system. extreme cases, the breathing circuit could be placed
under a vacuum, which in turn would put the patient’s
First a quick note on units of vacuum. In the patient lungs under a vacuum and could be fatal). If the APL
care environment, pressures and vacuums are expressed valve were to see excessive backpressure, the breathing
in millimeters of mercury (mmHgV) or inches of water circuit and the patient’s lungs could also experience
column (inH2OV). In medical vacuum design, we are higher pressures. The function of the interface valve to
used to expressing vacuum in inches of mercury (inHg). the WAGD system is to prevent these two hazards.
Perfect vacuum for those in the medical gas world is 30

WAGD Guide 7
 WAGD Interfaces

Flowmeters

Anaesthetic Agent
Wall Outlets Vaporizer

2 %
x-ANE

Fresh Gas Inlet

Adjustable Pressure Limiting Valve
(APL valve)
WAGD Connection
To Patient
Inhalation Exhalation
Valve Valve
P

Rebreathing Breathing
Bag
P =Monometer CO2 Absorber
Circuit
(Breathing Circuit
Pressure Guage)

Figure 2
The Anesthesia Machine

As we set about designing the WAGD systems to The WAGD Interface (Ref Figures 3 and 4)
connect to these interfaces, the above criteria is worth
restating, particularly because to engineers it may at first The interface between the anesthesia machine and the
appear counterintuitive. A properly functioning WAGD WAGD inlet has two major functions:
system (including the interface) must remove any waste
gas from the APL valve, while never permitting either a 1. It must guard the breathing circuit from exposure to
vacuum or a pressure to appear at the APL valve itself. vacuum or pressure. If the vacuum at the WAGD inlet
is 28 inches of mercury (as might be seen in a dual use
The pressure and rate of flow from the APL valve is quite WAGD/Medical implementation) the interface valve
variable. Pressures in the breathing circuit fluctuate must reduce that to a vacuum which will not disturb the
depending on the patient’s breathing. Naturally, the pressure relationship at the APL valve. In essence, the
pressure in the circuit will drop during inhalation and vacuum must be turned down by a factor of as much
rise during exhalation. This will be reflected in changes as 400. Naturally, a turndown this extreme is tricky to
in the flow at the APL valve. Patients might sigh, achieve and maintain, especially when one considers
hiccough, stop breathing, cough, and otherwise cause the variations in flow and pressure at the APL valve.
the pressures and flows at the APL valve to fluctuate
wildly during a procedure. Adults will produce more 2. It must ensure that the inevitable surges in waste
flow than infants. The diagram does not account for the gas volume are contained and passed into the WAGD
anesthesia ventilator, which of course will also vary the system rather than overflowing into the room.
pressure in the breathing circuit. The WAGD system
must be able to handle all of these variations without The most common North American interfaces are
dumping waste gas into the room. Most of this balancing diagrammed in Figures 3 and 4 (These diagrams are
act is performed by the WAGD interface. simplifications of the actual interfaces for ease of

8 WAGD Guide
WAGD Interfaces

Control Over-pressure From AP Valve

understanding).
Valve Relief
Figure 3 represents what
are termed “closed”
interfaces. To
WAGD
A tube from the APL Inlet
valve connects at
the interface inlet.
The interface outlet
is connected to the Under-pressure
WAGD inlet on the wall Relief
or ceiling. Control is Containment
achieved with a valve, Bag
which is intended to
balance the inflow
Figure 3
from the APL valve line
A “Closed” Interface
with the outflow to the
WAGD system. This
balance can be roughly Flows through these interfaces and into the WAGD
observed by the behavior of the containment bag - if system are typically 6-9 lpm when properly adjusted.
it inflates, the outflow is too small. If it deflates, the However, versions of these interfaces are now calibrated
outflow is too great. to draw 40 lpm when 12 inHgV is drawn at the interface
outlet (i.e. using a fixed orifice instead of a valve).
If more flow comes to the interface than it can handle,
(i.e. a surge from the breathing circuit or the needle European interfaces are based on another design
valve is not sufficiently open) the bag will inflate and sometimes called an open interface, a Bohringer tube
ultimately there is an overpressure relief which will or tube in a tube. These interfaces are simpler and
dump the excess gas into the room. If the flow from the operate on the most basic of physical principles, but are
breathing circuit is too small or the valve is open too far, nevertheless quite effective when properly used.
there is an underpressure relief which will admit room
air to prevent a vacuum in the interface valve. Again, simplified for easy understanding, the interface
works roughly as described in Figure 4.
However, the control valve is sometimes left wide open
and the underpressure relief valve(s) simply remains The tube from the APL valve connects at the interface
open. Although this is incorrect usage, the interface is inlet. The interface outlet is connected to the WAGD
so designed as to be able to handle this as long as the inlet on the wall or ceiling. Unlike the closed interface,
interface is properly maintained. these interfaces often do not have manual adjustments.
In those cases, a fixed orifice is simply placed in the
A criticism of this style of interface is that they rely outlet calibrated for the source vacuum and the required
too much on the underpressure relief valves. If the flow. Air is continuously drawn through the outer tube
underpressure valve becomes clogged with the lint or which completely encloses the inner tube. This constant
debris they can suction up when open, they may not flow and slight residual vacuum induces a flow in the
function to protect the patient. Current designs for these inner tube. The waste gas, pushed by the slight positive
valves specifically minimize this risk, for instance by pressure at the APL valve and pulled by the vacuum
providing redundant underpressure relief valves. passes through the inner tube where it is simply pulled

Flow Control Orifice From AP Valve

Outer Tube

To
WAGD Inner Tube
Inlet
Figure 4
An “Open” Interface D

WAGD Guide 9
 WAGD and NFPA 99

out through the orifice along with the continuous flow Of these five, the last often must be immediately
along the outer tube. disregarded. Handling WAGD through a passive system
has long been discredited as ineffective and in most
Surges in waste gas flow are accommodated by the HVAC arrangements cannot effectively comply with
excess gas simply flowing down the outer tube. The 6.4.1.6. The arrangement of a typical O.R. makes the
dimension “D” and the relative diameters of the tubes air very difficult to change and waste gases may linger in
are critical. Carefully selecting these dimensions limits the many dead spots created by the equipment, drapes,
the vacuum as well as providing capacity for handling etc. The air changes have to be set impracticably high
the expected surges without the waste gas escaping into and no recirculation is permissible, which makes such
the room. an approach undesirably expensive.

An underlying assumption for open interfaces is that the Venturi driven systems are unknown in North America,
flows will be large. High flow is necessary to prevent although they are not unusual elsewhere in the world.
spillage from the tube. Naturally, the higher the flow They offer the designer only limited advantages over
the smaller the interface itself. Flows of 80 to 130 liters other systems and are a relatively expensive and
per minute (2.8 to 4.6 scfm) are required under the complex option. We will discuss them under distributed
British Standard 6834:1987, and 25 lpm minimum, 50 producer systems.
lpm maximum is required under the newer European
Standard EN 737-2 1998. High vacuum systems (dual use and dedicated) are the
most common implementation in North America. They
The tubes are relatively large, which means the air use a vacuum pump as their source, typically running
velocity is low at the end. This in turn reduces the at the same vacuum levels as the medical vacuum (15
risk of sucking up debris which might block the tube. inHgV - 29 inHgV).
Some interfaces are provided with filters, and some with
indicators or flowmeters to let the operator know they The standard permits almost any implementation for a
are within specification. dual use high vacuum WAGD/Medical system (see Fig.
5 for examples).
It is important to note that while either interface type
could work with any WAGD system, the interface and Dedicated WAGD permits two different implementations:
the system must be matched in terms of pressure and high vacuum (pump driven) and low vacuum (blower
flowrate. This may mean adjusting orifices or other or fan driven). These are not very different in the
modifications to adapt the interface. requirements from the NFPA 99 standard, but require
very different engineering.
WAGD According to NFPA 99
When we consider the basic requirements of designing a
NFPA 99 has several sections on WAGD which are system we will revisit the specifics of these requirements
useful as a starting point for the design of the systems. as they apply to alarms, valves, etc.
The standard however is not comprehensive and the
designer must also have other sources to rely on. The Ideal WAGD System

NFPA permits five different WAGD implementations. What constitutes an “ideal” WAGD system? These
They are: characteristics apply to either WAGD interface and
might be used as something of a laundry list for anyone
1. Dual use WAGD/Medical vacuum (high vacuum). designing WAGD.

2. WAGD into a dedicated pump (high to moderate • The ideal system is active. That means there is some
vacuum). form of a “motivator” (NFPA uses the term “producer”)
which actively moves the waste gas down the line.
3. WAGD into a blower or fan (low vacuum) Passive systems (which rely primarily on the slight
positive pressure from the APL valve) are not effective
4. WAGD into an inlet driven by venturi. The venturi with the newer interfaces, and arguably never worked
must be driven by some system other than medical air adequately.
(instrument air is ideal).
• The ideal “producer” is oxygen inert. There is nothing
5. WAGD handled passively e.g. by ventilation and air in the producer likely to be a fuel for a fire.
changes (see Chapter 6, Environmental Systems, 6.4.1.6)

10 WAGD Guide
The Ideal System

• The ideal system operates at as low a vacuum level

What makes a pump “sensitive to oxygen” or
as possible. There is nothing more essential to patient
safety than preventing the breathing circuit from seeing “suitable for use with WAGD”?
a vacuum. The lower the maximum system vacuum, the
Vacuum pumps vary greatly in their suitability for use
lower the risk, without having to rely on the interface.
with WAGD, based primarily on their compatibility
with oxygen (although there are other considerations,
• The ideal system has a high flow. At a minimum, each
they pale compared to this one).
inlet must be able to draw a continuous 50 lpm (1.8
scfm) through the interface. The most important consideration is simply the
availability within the pump of fuel, which practically
In light of these criteria, there are two questions to be means oil or grease. Any pump in which oil or grease
answered: are in contact with the gases is therefore unsuitable.
However, to be acceptable for WAGD does not
1. What do I do if I have an existing WAGD system and mean to be suitable for oxygen. As an example,
am going to attach new anesthesia systems to it? a vacuum pump for WAGD does not need to be
“oxygen clean” in the way an oxygen pipeline needs
2. How do I design any new WAGD systems? to be clean, or the way the same pump would need
to be cleaned if it were actually pumping oxygen.
Dealing with an Existing System Very small amounts of oil or grease, such as might
be found on a machined part are unlikely to cause
What do I do if I have an existing WAGD system and am problems in this service. The greatest concern arises
going to attach a new anesthesia system to it? not simply from the mere presence of oil, but the
presence of oil in quantity.
Each of the many combinations of system layouts,
producer technologies and other variables implies it’s Similarly, graphite vane pumps are not ideal for
own risk profile and will result in a different priority WAGD service, because the vane dust can form a
order for the available options. fuel and is present in some quantity. Nevertheless it
needs to be said that we have no reports of fires with
As discussed, there are three primary concerns to graphite pumps in WAGD service.
consider. The most immediate is the oxygen sensitivity
of the existing system. It is clear that any possibility of In rough order of concern, common pump
fire is the greatest hazard. A less critical issue, but one technologies fall out roughly as follows:
equally demanding a solution, is the problem of higher
inflows. The third is the patient safety concern involved Oil lubricated: These are typically rotary vane, oil
with any high vacuum implementation (see page 7). seal liquid ring, screw or reciprocating pumps which
contain large volumes of oil. These are at great risk
Figure 5 illustrates a decision process for considering with elevated oxygen. All reported fires have been
each of these concerns in order. Reference to the figure in this category of pump.
will lead the reader through the necessary decisions,
explained under the main headings below: Dry or graphite vane pumps: These are rotary vane
pumps which use no oil in the machine but may have
WAGD Inlets: lubricated bearings and use sacrifical graphite vanes.
It is a basic requirement that any facility must have These are probably moderate to low risk.
dedicated WAGD inlets distinct from the vacuum inlets,
Regenerative blower (dynamic) and dry rotary lobe
marked “WAGD” or “Evacuation”, colored purple and
pumps: These machines have air ends in which there
not interchangeable with the vacuum inlets. These have
is no source of fuel (i.e. no oil). Although there may
been required under NFPA 99 since at least the 1996
be lubricants in their bearings and in their gear cases,
edition (ref NFPA 99 1999 edition, 4-3.3.2.3 (b).) If the
the oil containing chamber(s) and the compression
facility does not have such dedicated WAGD inlets,
chamber are separated. Generally, these machines
installation of a new WAGD system will be required.
are low risk as long as they are well maintained.
Dual use vs. Dedicated WAGD Producers: Water Seal Liquid Ring: These pumps contain no oil
The hazards involved with Dual Use and dedicated nor do they allow any practical way for the oxygen to
implementations are nearly identical if they are both reach their bearings. They also run very cool. They
driven with pumps, but removing any oxygen sensitivity present the smallest risk.
will involve very different levels of complexity and cost.

WAGD Guide 11
 Existing Systems

Dedicated systems will of course also have dedicated problematic. The most commonly manifestation is
piping, which greatly facilitates any solution. simply inadequate capacity. In any case where a pump
is stressed (i.e. running too much or calling in the lag
High Vacuum and Patient Safety: pump), and the WAGD is demonstrably the cause, two
The concerns over the patient safety issues with any high basic options must be considered: dividing the systems
vacuum implementation are explained on page 7. (page 13) or replacing and upsizing the pump (below).

Oxygen sensitivity in producers: On the use of Medical Air to run the anesthesia ventilator:
See the sidebar (page 11). If the O.R. has piped medical air, it may be possible
to have the anesthesia machine modified to run the
Pump capacity: ventilator on medical air rather than oxygen. While
A pump may be compatible with oxygen but still be this does not eliminate the concern with flowrate, it

�����

������
��������� ���
�����������
��

������� ��������������
����������� ��� ��������� �� ������������ ���
����������� �������� ������
������������
������������

����������������
��
�
������������������
�������������� ���
�����������
�� ������������ �������������� ��
��������
����������
����������� ��� ������������ ��
������ ����������
��������������
��������������
���������� ��� ���������
������������
�������
�������������� ���
�� ��������
�������������������
�������������������
������������������ �������
�������� ����������������� ������ ���������������
�����������
����������
�����������
�����������
��������������
��������������
������������
�����

Figure 5 ����������������������������
A Decision Tree for Existing Systems ��������������������������������������

12 WAGD Guide
Replacing Producers

does greatly reduce any hazard of fire in the pump. This These specialty lubricants often outlast standard oils, but
can only be considered if the O.R. has piped air, since they still need to be changed periodically. Whenever
the volumes required will rapidly deplete cylinders. the lubricant is changed, it must be done with this
Medical air is not required in the O.R. under the 2001 special lubricant both to preserve the oxygen compatible
AIA Guidelines for Design and Construction of Hospital character of the pump and because the standard and
and Health Care Facilities. To investigate air as an special lubricants cannot be mixed.
option, you must have the answer to three questions:
This kind of a conversion should only be undertaken with
1. Is medical air piped into every location anesthesia full knowledge of the total costs and with a complete
machines will be used? include in your survey understanding of the manufacturer’s limitations on the
outpatient, inpatient, delivery, trauma and any other use of the converted pump.
specialty O.R.
The comments above are primarily directed at oil
2. Is there sufficient medical air capacity? For lubricated technologies such as liquid ring oilseal,
calculation, you should assume 2 scfm per O.R. with screw and lubricated rotary vane. It is possible and
a 100% simultaneous use. even desirable to consider this kind of adaptation for
technologies which are oil free as a way to reduce any
3. Is the medical air in good condition? It is still true remaining risk. For instance, in rotary lobe machines
in some facilties that medical air is wet or dirty. In and in dynamic machines it may be possible to use
intensive care, it is often found that the medical staff these oxygen compatible lubricants with only a small
has abandoned the piped air for these reasons, using increase in operating cost because of the longevity of the
instead portable compressors for their ventilators. lubricant. Again, the manufacturer must be consulted
This is not an option in the O.R., so the piped medical before taking any action of this type.
air must be as required by NFPA 99 before it can be
used. Second: The pump may be replaced with a pump of
appropriate capacity but which is inherently oxygen
Replacing a Pump or Producer compatible. When this is done, consideration must be
given to the additional WAGD demand. In many cases,
Replacing a pump or producer is often the quickest, it may be necessary to use a larger pump. The comment
surest and cheapest solution. Obviously, if capacity in the paragraph above on the use of oxygen compatible
is the primary problem, then replacement offers an lubricants is applicable to this approach as well.
immediate solution. If oxygen sensitivity is the problem,
replacement alone can resolve this issue without further It is wise to recheck pipe and exhaust sizing, electrical
changes to the systems as piped. capacity and ventilation, especially where a pump is
upsized.
When considering pump replacement, there are a
number of considerations. Dividing an Existing Dual Use System

First: In some cases it is possible to change not the pump When a decision to divide the systems is taken, there
but the lubricant/sealant within the pump. There are are many, many variations in how this might be
halogenated lubricants which can withstand contact accomplished. In order to evaluate a decision, you will
with oxygen and can be used in some existing pumps need the answers for the following:
to make them “oxygen compatible”. Note that these
lubricants are NOT simply “synthetic oils” (which are 1. How is your present system organized? Figure 6
essentially the same as natural oils as far as oxygen illustrates the common layouts. Systems resembling
sensitivity is concerned). arrangement “A” are very easily and inexpensively
divided, and lend themselves best to a dedicated
Pursuing this option must first be discussed with the pump. Systems resembling “B” are difficult to divide,
supplier of the pump to ensure it is even possible. The but allow a number of options in doing so. In systems
conversion typically requires that the pump be largely or resembling “C”, the options are many but they all
totally disassembled, completely cleaned, reassembled, have complications. There will have to be significant
and filled with the special oxygen inert lubricant. In most repiping, possibly inside ceilings or walls and even
cases the pump must be returned to the factory. Costs inside the ceiling columns, pendants, etc., depending
will include preparation and the first fill of lubricant. where the WAGD unites with the vacuum piping.
The whole process can be grotesquely expensive and
the replacement lubricant is amazingly costly. 2. Is the objective a complete update of the WAGD system

WAGD Guide 13
 WAGD Implementations

Sample Arrangement A An Introduction to WAGD System

Mixed System Separate Piping, Implementations
Joined at source
We will now examine in detail four of the
five acceptable WAGD implementations
under NFPA 99, along with some
alternative ways they may be implemented.
As mentioned previously, we will not deal
with Passive implementations.

Dual use Medical Vacuum/WAGD

implementations:
Sample Arrangement B
Mixed System, Joined in line Dual use implementations are primarily
medical vacuum systems into which the
WAGD is introduced, more or less as a
“free rider” on the medical vacuum network
and pump.

Dual use implementations are high vacuum.

The Medical vacuum application will
always take precedence when setting pump
Sample Arrangement C
cut in and cut out. This means the lowest
Mixed System, Joined at inlet
vacuum a WAGD inlet can see is 12 inHgV
(305 mmHgV) (per NFPA requirements for
medical vacuum) and typical vacuums are
15-29 inHgV (380-760 mmHgV)

Due to the high vacuum, the networks for

these systems must be relatively strong.
Copper pipe is commonly used.

Figure 6 WAGD terminals look very similar to

Dual Use Implementations vacuum inlets except for color, and are
nearly identical to vacuum inlets internally.
Most significantly, this means that the
to minimize all risks or merely enough to attenuate an internal porting is limited, and many
immediate fire risk? A complete update will imply WAGD inlets (particularly older ones) may have some
changing to a low vacuum methodology (see below difficulty with higher flows.
“An Introduction to WAGD System Implementations).
An overhaul to reduce immediate fire risk may be Piping is typically sized using the same loss tables
much less sweeping. It is of course also possible to do as are used for vacuum, and valving and alarming
some of both in many circumstances. generally follows the vacuum rules.

3. Are there obstacles outside of the engineering and A rare variant on this implementation, which can be
construction related challenges? Is the anesthesia employed most simply where the system layout is like
department aware of the issues and supportive of the Figure 6, Arrangement “A”, is a regulated WAGD line.
improvements? Can they adapt their older anesthesia In these implementations, a regulator is installed in the
equipment to operate acceptably with the proposed WAGD line which reduces the top vacuum level to a
new WAGD system? medium or low vacuum and thus brings the WAGD
pipeline closer to the ideal of an intrinsically safe
Whether for an existing system upfit or a brand new system. Such an implementation requires separate
system, it is important to understand the system options sizing of the WAGD line at the lower vacuum.
and their respective engineering challenges. Copper pipe would still be the preferred material of
construction.

14 WAGD Guide
WAGD Implementations

Dedicated WAGD, using a Pump: Unlike the two types discussed previously, these systems
typically use a regenerative blower as their producer.
Pump-based implementations have in the past generally Whereas a pump is designed to expand the air first (i.e.
been run at vacuum levels similar to medical vacuum. produce a deep vacuum) and move volume second, a
Pumps usually selected are designed to run at vacuums blower is designed to move volume first and produce
of 15 inHgV {380 mmHgV} or higher. However, a only a shallow vacuum. In this they more closely
variant of this system is a medium to low vacuum system resemble a fan than a pump. It is this emphasis on
wherein the systems operate more nearly in line with the moving lots of air at low vacuums that makes them ideal
ideal of an intrinsically safe system. This is achieved by for WAGD.
sizing and setting the pumps to lower vacuum settings
(e.g. 5-10 inHgV {127-250 mmHgV}) and sometimes As an example, a liquid ring pump with a one Hp. motor
also installing a vacuum regulator in line to reduce the will move 396 lpm (14 scfm) at 5 inHgV (if it could be
vacuum at the inlet. This operated there - most will
variant is appropriate only A Remote Source run up to 28 inHgV). A
with technologies which regenerative blower driven
are suited to low vacuum with the same 1 Hp. motor
operation (liquid ring, dry will move 1,245 lpm (44
vane, rotary lobe). cfm) at 40 inH2OV (2.9
O.R. A O.R. B inHgV).
The networks must be
designed to handle the full This low vacuum creates it’s
vacuum of which the pump own issues. The sizing of
is capable (even though the O.R. C O.R. D the network is immediately
pump may in fact operate at different from that used
a lower vacuum normally). Remote for higher vacuum, and
Therefore, the networks Mechanical Producer the terminals must be of a
Room
for these systems must be different type to pull enough
relatively strong. Copper flow at these low vacuums.
pipe is most commonly A Local Source
used. However, an important
Local Producer advantage of these low
At vacuums 12 inHgV {305 vacuum implementations
mmHgV} and above, piping is that the producer may
would be sized using the be located close to the
same loss tables as are used terminals, unlike pumps
O.R. A O.R. B
for medical vacuum, but which typically must be
alternative sizing methods Local Producer located remotely. A blower
will be required at lower (especially a small one) is
vacuums. sufficiently compact and
Figure 7 quiet that it can sometimes
O.R. C O.R. D
WAGD terminals used at the Remote vs. Local Producers be placed near the WAGD
upper end of the vacuum terminals (e.g. in a ceiling
scale (5 inHgV and higher) space, a mop closet, etc.)
look very similar to vacuum inlets except for color, and The network can therefore be minimized. A pump,
are nearly identical to vacuum inlets internally. Most being typically larger and more noisy must typically
significantly, this means that the internal porting is be placed at some remove in a mechanical space.
limited, and many WAGD inlets (particularly older ones) The resulting network is longer and potentially more
may have some difficulty with higher flows. Therefore, complex. (see Figure 7)
at the lower end of the vacuum scale the terminals may
need to be of a different type to ensure adequate flow. At first glance, it may appear to be more expensive
to implement and operate a local system possibly
Dedicated WAGD, using a Blower: composed of multiple blowers vs. a single large pump.
In fact, because blowers are less expensive, smaller,
These implementations achieve all the objectives desired and internally less complex than pumps of similar
in a WAGD implementation, but typically are the most capacities, the economics often slant in favor of the local
complex to design and install. implementation.

WAGD Guide 15
 WAGD Selection

Low vacuum systems can be piped in a variety of 1. Effectiveness. Will the system do the job of keeping
materials. NFPA requires they be metallic and non- the workspace free of waste gas? Efficacy usually has
corroding, which rules out plastic or iron pipe. However, less to do with the type of system selected than the design
copper, stainless, and galvanized pipe might be used, as and installation of the system, as all the implementations
might ductwork and thin wall galvanized because of described herein are perfectly capable of being effective
the low vacuum (an ideal material would seem to be if well designed.
electrical conduit with liquid tight fittings, but so simple
an answer may be too exotic to be acceptable in many 2. Patient safety. Will the system protect the patient and
local jurisdictions). Copper pipe is the most common ensure the anesthesiologist’s control of the procedure?
material. Here, the low vacuum implementations are to be
preferred over the high vacuum implementations due to
These systems may require a means to balance the the intrinsic safety implied in a lower vacuum.
system and are tested in a somewhat different manner to
high vacuum systems. 3. Cost. Which system is least expensive to implement
and operate? Evaluating this is complex and the result
WAGD using Distributed Producers: varies dramatically between facilities. In general, low
vacuum systems are less expensive than are high vacuum
These implementations are in many ways the simplest, systems. Low vacuum systems are also typically lower
but have not been seen in the North American market maintenance than are high vacuum systems.
and are thus unfamiliar. They tend to be relatively
expensive to install because two separate piping However, an assertion often made in favor of dual use
networks are required. implementations is that WAGD dumped into a medical
vacuum system is “free”, since the medical vacuum “has
Conceptually, they involve no central producer but to be there anyway”. When the average WAGD inlet
instead use a venturi actually in each inlet. The venturi only flowed 6-9 liters, there were many cases where this
is intrinsically capable only of low vacuum. was at least in part true. With WAGD flowing at 50
liters (1.8 scfm), it is true far less often. The additional
The venturi must be served by an air line, and medical capacity required and the additional operating hours
air is not permitted to be used for this service. Instrument mean that the cost of WAGD produced by a medical
air is ideal, but at this writing relatively few facilities vacuum pump is considerably higher than has been
have instrument air systems. Once the venturi is served assumed.
by an appropriate air source, the exhaust must also be
routed to the outside and sized. A simple rule of thumb test can be applied: Size the
medical vacuum system without WAGD and select
The exhaust side may be made of pipe suitable to the a pump of appropriate capacity. Add in the WAGD
pressure. Copper would be typical. There are no requirement (use at least 1.8 scfm per location). If
alarms which can practically be installed. Each inlet the pump selected has sufficient capacity to handle
is individually controlled, and must have an operating the additional volume, an argument can be made that
indicator of it’s own. the WAGD produced is “free” or at least low cost. If
the pump selected does not have the necessary excess
All WAGD implementations except the distributed capacity, and thus to accommodate the WAGD a larger
producer styles share the same basic requirements as pump must be selected, a properly selected dedicated
to the location of terminals, alarms, etc. All WAGD system will almost certainly be less expensive. This is
implementations share the same basic requirements as especially true when a low vacuum system is used for
to the discharge from the building. Specific details of comparison. Remember that a horsepower of pump will
how these items operate will vary by implementation. move approximately 15 scfm, whereas a horsepower
We will consider the universal requirements here and of blower will move approximately 44 scfm, a ratio of
then deal with the necessary variations under each roughly 3:1.
specific implementation.
4. Technology. Is the technology otherwise preferred
WAGD Selection for the medical vacuum source acceptable for WAGD?
If not, can another option be equally acceptable? In
What factors should be weighed when selecting a some cases a technology otherwise preferred for use
WAGD implementation? There are several, and the with medical vacuum may be oxygen sensitive, and
weighting to be given to each will vary from facility to there is not an equally acceptable oxygen compatible
facility. They include: alternative. Naturally this will restrict the potential

16 WAGD Guide
General Requirements

�������

���� �� ����

����
����
������
���������
������
������� ����� ��������
���� ������ ����� ���
�����
����� �����
�����
���� ����
����� �����
����
��������

����
������

�������
��� ��������� ����� ������
����� ���

���� ����
����� �����
������
Figure 8 ����� ���
General Requirements for any Dedicated WAGD Implementation
(diagram is for general illustration only - requirements for specific
occupancies vary)

for dual use. The same limitation may also render location piped with nitrous oxide. Consideration should
unacceptable pumps otherwise preferred in a dedicated also be given to areas which are not traditionally piped
pumped system. In such a case, the limited technology with nitrous, but where nitrous oxide mounted on the
options may be a powerful argument in favor of a low anesthetizing machine can reasonably be expected to
vacuum alternative. be used. Common examples of these locations are
CAT/MRI/PET scan rooms, lithotripsy, exam/treatment
5. Design complexity. How difficult is the system to rooms, trauma rooms and other locations where patients
design and what are the chances of problems resulting are routinely sedated or anesthetized. Other areas
from bad design? Whatever can be said against a dual which should receive consideration might include
use system, they are undoubtedly among the simplest laboratories where veterinary anesthesia will occur,
to design. Correspondingly, low vacuum dedicated dental clinics and OB/Labor where nitrous oxide/oxygen
systems offer the greatest range of advantages for the self administration is practiced. WAGD terminals may
user, but are probably the most complex to design and also need to be placed in areas such as recovery, where
are also outside the experience of most North American exhaled anesthesia from recovering patients must be
designers. Low vacuum dedicated systems are also the considered and the staff protected.
most complex to commission.
All dedicated WAGD producers are required to be
WAGD Design : General Requirements duplex wherein one unit must be sufficient to serve the
(Reference Figure 8) system, and a second of equal size is ready to operate in
the event of any inadequacy in the first. A local alarm
NFPA 99 states that a unique, dedicated WAGD indicating Lag WAGD producer in service must be
terminal should be placed wherever nitrous oxide or included and relayed to the master alarm. The WAGD
halogenated anesthetic is intended to be administered producer is required to include a source valve.
(13.3.5.2, 14.3.5.2). This will obviously include any

WAGD Guide 17
 Common Questions

Exhausts from the producer must exit the building, trace gas tests or struggling to pass, then action must be
discharging at least 10 feet from any opening in the taken to make the system work better and you can follow
building, at a vertical level different from any air intake the guidance in this pamphlet to evaluate your options.
(preferably at a lower level) and in the a location which Notwithstanding how you make out on the trace gas tests,
is open and permits free dispersion of the waste gas. check for excessive run time at the vacuum pump. If
your vacuum pump is running hard or can’t keep up with
Electrical power must be from the essential electrical demand, this may be because the WAGD requirements
system, equipment branch. have begun to run away.

Centrally piped WAGD is required to be valved like Q. Should I be concerned if my present WAGD pump
any other medical gas or vacuum system. Valves contains oil?
may be either ball or butterfly type. Required valves
include: Source valve, main valve (in a limited number A. You should always be concerned if your WAGD goes
of circumstances), riser valves, service valves and zone through a pump containing oil. It is very strongly advised
valves. that you take some action to mitigate the possibility of oil
and oxygen mixing in such pumps.
WAGD inlets must be separate from and non-
interchangeable with the vacuum inlets (even if they Q. We’re about to get some new anesthesia machines.
ultimately are piped to the same source). WAGD Can I pretest my WAGD?
has it’s own color code (White letters on violet).
Many older WAGD or Evacuation inlets were “one A. Yes. If your WAGD goes through a pump containing
way” interchangeable (you could plug vacuum into oil you should immediately look to make changes. With
Evacuation, but not Evacuation into vacuum). This is no other pumps, you can check the pump’s spare capacity by
longer permitted and such inlets should be retrofitted to simply timing it’s operating cycle. In addition, you must
bring them up to standard. test each WAGD inlet for inflow capacity of at least 50
lpm (1.8 scfm). The test in Annex A is the test required in
The master alarms for a piped WAGD system will the EN 737-2 standard, and will be useful in determining
include at least an indicator for “Low WAGD” and an if your current inlets operate in the required flow range.
indicator for “WAGD Lag Producer Running”
Q. We run our ICU ventilators on air, can’t we do that
Any area fitted with piped WAGD requires a WAGD area with the anesthesia ventilators?
alarm at the nurses station just like any other medical
gas. It will typically be piped into the line upstream of A. Yes, if medical air is piped to all your O.R.s and has
the anesthetizing location zone valve. sufficient capacity, the anesthesia machines can be
modified to run off medical air. (see page 12)
Distributed WAGD may be considered exempt from
some of these requirements, but others may need to be Q. Does pipe and fittings used for WAGD service need to
fulfilled in unusual ways. As an example, while it is be cleaned for oxygen service?
impractical to alarm a venturi, it is appropriate to alarm
the drive air so that the facility knows if the WAGD is A. No. They must be treated according to the same rules
inoperable. It is necessary to observe intent in these as apply for vacuum. However, use of cleaned pipe and
cases and to ensure the essential functions are present fittings is always to be recommended.
even if it is necessary to use different methods to achieve
the result. Q. Does all my WAGD have to be on a single system?

Common Questions A. No. In fact, with multiple areas of the facility being
involved, it is often easier and less expensive to implement
Q. How do I know if my current WAGD system is a separate, purely local WAGD system for each area.
O.K.? This same principle applies with distributed WAGD.

A. You should always be concerned if your WAGD goes Q. I have an older WAGD system implemented through
through a pump containing oil (if it does, see below). ductwork and run with a fan. Do I have to replace it?
With any other type of pump the most important test is the
periodic test for trace gases in the O.R. If you are passing A. Not if it is working (i.e. as demonstrated by passing
this test without any problem, then the chances are very periodic trace gas tests). These “semi passive” systems
good your system is doing it’s job. If you are failing the are sometimes quite adequate and there is certainly no

18 WAGD Guide
Sources & Annex A - Testing

reason to replace a working system. If your trace gas tests References

are coming up clean, then the system can be continued in
use without concern. If you find the tests are getting hard American Institute of Architects Academy of
to pass, then you will have to consider an active WAGD. Architecture for Health
Be aware that some of the newer anesthesia systems do Facilties Guidelines Institute
need “pull” (i.e. a slight vacuum) at the WAGD inlet, and 1919 McKinney Ave.
that these semi-passive systems may not have enough Dallas, TX 75201
“pull” to be effective with EN compliant interfaces. www.aia.org

Q. Can I use plastic pipe for WAGD? BeaconMedæs

www.beaconmedaes.com
A. There is no problem with plastic in terms of the gases
or sizing, but plastic pipe is not allowed by NFPA 99 or British Standards Organization
most local jurisdictions in commercial occupancies. 389 Chiswick high Road
London W4 4AL United Kingdom
Q. With a change to my WAGD system do I have to re- www.bsi.org.uk
verify?
European Committee for Standardization (CEN)
A. Yes. A change to a WAGD system needs to be handled Rue de Stassart 36
like a change to any medical gas system. B-1050 Brussels, Belgium
www.cenorm.be

Medical Gas Professional Healthcare Organization

(The Scott study cited in this paper (p. 6) will be found
on this website in the “Forum” section.)
www.mgpho.org

National Fire Protection Association

Batterymarch Park
Quincy, MA 02269
www.NFPA.org

To send questions, comments or suggestions,

or to obtain additional copies of this pamphlet:
mallen@beaconmedaes.com

WAGD Guide 19
 Annex A - Testing

Annex A Probe
matching Inlet
Testing a WAGD Inlet (may have to be connected

kPaV
to the test apparatus through
a hose)
The EN 737-2 standard gives a very simple and effective
method for the testing of an inlet and also for the testing Tee
of the WAGD system. The device required is illustrated Vacuum Guage
in Figure 9. or manometer

Control valve
The procedure is very simple and might be applied to “A”
or fixed orifice
any WAGD terminal. You may expect some odd results
when applying this test to standard WAGD inlets, as the �
EN Standard assumes a low vacuum implementation is in
���
use. Nevertheless, it is possible for a high vacuum imple-
�
mentation to function within the parameters of the test. 60
���
� 55
The test is conducted at very low levels of vacuum (1kPa
and 2 kPa {4 inH2OV - 7.9 in H2OV or 0.3 inHgV - 0.6 ��� 50
inHgV). If you intend to test a high vacuum inlet, be aware �

��������������������������
45
that the gauge used in this test is extremely precise and ��
probably delicate. It may therefore be ruined if exposed � 40
to high vacuums. Check the specification for the gauge Flowmeter 0-50 lpm
�� 35
you intend to use prior to using it and ensure it can handle
the maximum possible vacuum in the system as well as � 30
be accurate in the range required by the test. ���
25
�
The test apparatus consists of: ��� 20
� 15
1. A flowmeter which can measure at least between 25 ���
and 50 lpm. An appropriate device might be a 0-60 lpm 10 Figure 9
� Test Apparatus for the
rotometer open at one end and having a threaded outlet
5 EN 737-2 Flow Test
at the other.
(diagram is for general
2. A metering valve with as wide an opening (a high CV) illustration only - other
as is available. It is not particularly important that the arrangements are
valve be capable of highly precise control but it is essential possible)
that the valve not make the test impossible to perform by
too great an internal resistance. It is also possible to do When a test reaches one of the stop signs in Figure 10,
the test very accurately with fixed orifices rather than a it helps to understand that the EN views the inlet as in
valve if the vacuum in the system is known and does not essence an orifice. The vacuum on the pipeline side
fluctuate widely. is (or should be) known and relatively stable (this is an
important point, as the very large swing in vacuum level
3. A vacuum gauge accurate at 1kPa and 2 kPa {4 common in many dual use and pumped dedicated systems
inH2OV and 7.9 in H2OV or 0.3 inHgV and 0.6 inHgV}. may alone be the reason they may not pass the test.) The
This may be an actual water column in a low vacuum inlet itself can then act as a controlling orifice, and inlets
test. designed to the EN standard often actually physically
contain an orifice.
4. A probe or adapter matching the inlets.
As an example, in a low vacuum system, the vacuum level
5. Hardware as needed to assemble the components as in the piping is controlled to 5 inHg. This inlet contains
shown. It is important to minimize the internal resistances an orifice calculated to admit 50 lpm at 5 inHgV. When
in the test device, so selection of the components should the flow test instrument is in place, a “drag” is placed on
be made with care. Larger fittings are always better. the inlet of 1 kPa or 4 inches of water column, and the
inlet must flow 50 lpm or less. Increasing the “drag” to
To test a WAGD inlet, follow the decision tree in Figure 7.9 inches of water column, the inlet must flow greater
10. than 25 lpm.

20 WAGD Guide
Annex A - Testing

In a system like the example above, when an inlet has too little flow, it is relatively simple to enlarge the orifice
in the inlet and increase the flow. If the flow is too great, a smaller orifice can be substituted. Since the system is
consistent in vacuum level, the adjustments are very straightforward.

When the vacuum level in the system is not consistent, these adjustments become
extremely challenging, involving the balancing of the inlet performance at the lowest
�����
vacuum and at the highest vacuum. Thus every inlet must be tested at two vacuums
and a compromise orifice fitted. If the variation in vacuum is too great, it may be
impossible to keep the inlet within the flow requirements of the EN standard at all.
��������������
�������������� With standard U.S. style inlets (designed for high vacuum service), the internal port-
�������������� ing of the inlet will also act as an orifice (albeit a rather inconsistent one). If testing
reveals an inlet which cannot provide sufficient flow, then it may be possible to
increase the internal porting in the inlet and thus effectively open up the controlling
orifice. Conversely, if an inlet exceeds the permissible flow, it may be possible to
���������������
restrict the porting in the inlet, thus reducing the flow. It may also be possible (except
��������������������
in dual use implementations) to reduce or increase the vacuum level in the system,
which will also influence inlet flow. In extreme cases, it may be necessary to add or
change an orifice in the probe or interface, which will require the assistance of the
anesthesia machine manufacturer. At least one manufacturer of interfaces includes
���������
����
a small flowmeter with an adjustable valve which acts to gate the flow
from the interface when using a high vacuum inlet. This is another
means of dealing with the control of the flow from the interface which
could be implemented with any high vacuum system.
�����
�������
������� There is an additional test which is not required by the EN
�������
�������
Standard but is highly recommended. This is the system test.
In a system test, all inlets are opened to a calibrated flow
and then each inlet is tested as described previously. The
������������������� result assures that each inlet conforms to the flow specifica-
������� ���������������������
�������
tion when the entire WAGD system in operation. This test
��������������
������ will identify design flaws and piping problems that would
ordinarily never be seen when testing only inlets.

The system test simply requires a calibrated probe be inserted

��������� in each inlet while the inlet flow
����
test is conducted. Each probe
is calibrated to flow 25-50
lpm, placing the entire
������������������� design load on the
������� ���������������������
��������� ����������������
system.
������
���������
����
������� �����
������� �������
��������
������� �����
������� �������
�������

�������
��������� ����
Figure 10
Performing and Interpreting the EN 737 Flow Test

WAGD Guide 21
22 WAGD Guide
 TM

13325A Carowinds Blvd • Charlotte, NC 28273 • Phone 1 888 4 MED GAS • Fax 1 704 588 4949
www.beaconmedæs.com