Anesthesia Vaporizers

Measurement of Vapor Pressure and Saturated Vapor Pressure

The following description is intended to provide an understanding of how, in principle, the SVP of a potent inhaled volatile anesthetic agent could be measured in a simple laboratory experiment. It will also help to conceptualize the pressure that a vapor can exert. Fig. 3.1A shows a simple (Fortin) barometer, which is essentially a long glass test tube that is filled with mercury and then inverted to stand vertically with its mouth immersed in a trough containing mercury. When the barometer tube is first made vertical, the mercury column in the tube falls to a certain level, leaving a vacuum (the Torricellian vacuum) above the mercury meniscus. In this system, the pressure at the surface of the mercury in the trough is that due to the atmosphere. In a communicating system of liquids, the pressures at any given depth are equal, therefore the pressure at the surface of the mercury in the trough is equal to the pressure exerted by the column of mercury in the vertical tube. In this example, atmospheric pressure is said to be equivalent to 760 mm Hg because this is the height of the column of mercury in the barometer tube.

In Fig. 3.1B , isoflurane liquid is introduced at the bottom of the mercury column and (being less dense than mercury) rises to the top, where it then evaporates into the space created by the Torricellian vacuum. The isoflurane vapor exerts a pressure, causing a decrease in the height of the mercury column by an amount equal to the vapor pressure that it exerts. If one continues to add liquid isoflurane until a small amount of liquid remains unevaporated on the top of the mercury meniscus ( Fig. 3.1C ), the space above the column must be fully saturated with vapor, and the pressure now exerted by the vapor is the SVP of isoflurane at that temperature. Adding more liquid isoflurane will not affect the vapor pressure as long as temperature remains constant. If this experiment is repeated at different temperatures, a graph can be constructed that plots SVP on the y -axis against temperature on the x -axis. Such curves for some of the potent inhaled volatile anesthetic agents are shown in Fig. 3.2 . Contemporary technologies for measuring the partial pressures or SVPs of gases and vapors are described in Chapter 8 .

Boiling Point

The SVP exerted by the vapor phase of a potent inhaled volatile agent is a physical property of that agent and depends only on the agent and the ambient temperature. The temperature at which SVP becomes equal to ambient (atmospheric) pressure and at which all of the liquid agent changes to the gas phase (i.e., evaporates) is the boiling point of that liquid. Water boils at 100°C at one atmosphere pressure because at 100°C the SVP of water is 760 mm Hg. The most volatile of the anesthetic agents are those with the highest SVPs at room temperature. At any given temperature, these agents also have the lowest boiling points (e.g., desflurane and diethyl ether boil at 22.9°C and 35°C, respectively, at an ambient pressure of 760 mm Hg). Boiling point decreases with decreasing ambient barometric pressure, such as occurs at high altitude.

Units of Vapor Concentration

The presence of anesthetic vapor may be quantified either in (1) absolute terms, expressed in mm Hg [or kilopascals (kPa)], or in mg per liter; or (2) volumes percent (vol%) of the total atmosphere (i.e., volumes of vapor per 100 volumes of total gas). From Dalton’s law of partial pressures (see the following section), volumes percent can be calculated as the fractional partial pressure of the agent; that is,

Dalton’s Law of Partial Pressures

Dalton’s law states that the pressure exerted by a mixture of gases (or gases and vapors) enclosed in a given space (such as a container) is equal to the sum of the pressures that each gas or vapor would exert if it alone occupied that given space (or container). A gas or vapor exerts its pressure independently of the pressure of the other gases present. For example, in a container of dry air at one atmosphere pressure (760 mm Hg) with oxygen representing 21% of all gases present, the pressure exerted by the oxygen (i.e., its partial pressure) is 21% × 760, or 159.6 mm Hg. Consider the same air at a pressure of 760 mm Hg but now fully saturated with water vapor at 37°C (normal body temperature). Because vapor pressure depends on temperature, the SVP of water at 37°C is 47 mm Hg. The pressure due to oxygen is therefore now 21% of 713 (i.e., 760−47) mm Hg. The partial pressure of oxygen is therefore 149.7 mm Hg.

If the concentration of an anesthetic agent in a gas mixture is known in volumes percent, it may be converted to the mg/liter equivalent using the following formula (see appendix for derivation) :

• where W = concentration in mg/liter

• C = concentration in volumes percent

• MW = molecular weight of the anesthetic agent

• F = temperature-barometric factor

• Where F = (760/p) × [1+(t−20)/273]

• where t = temperature in °C

• p = barometric pressure in mm Hg.

For example, to express 2% sevoflurane in units of mg per liter at 1 atmosphere pressure and a temperature of 20°C:

Note that volumes percent expresses the relative ratio or proportion (%) of gas molecules in a mixture, whereas partial pressure (mm Hg or kPa) or mg per liter represents an absolute value. Anesthetic uptake and potency are related directly to partial pressure or mg per liter and only indirectly to volumes percent. This distinction will become more apparent when hyperbaric and hypobaric conditions are considered.

Minimum Alveolar Concentration

The minimum alveolar concentration (MAC) of a potent inhaled anesthetic agent is the concentration that produces immobility in 50% of patients undergoing a standard surgical stimulus. ^, Used as a measure of anesthetic potency or depth, MAC is commonly expressed as volumes percent of alveolar (end-tidal) gas at one atmosphere pressure at sea level (i.e., 760 mm Hg). Table 3.1 shows how MAC expressed in familiar volumes percent can be expressed as a partial pressure in mm Hg. Anesthesiologists should learn to think of MAC in terms of partial pressure rather than in terms of volumes percent because it is the partial pressure (tension) or concentration (mg/liter) of the anesthetic in the central nervous system that determines the depth of anesthesia. This concept has been advocated by Fink, who proposed the term minimum alveolar pressure (MAP), and James and White, who suggested minimum alveolar partial pressure (MAPP). In a 2015 editorial, James et al. proposed that “all anesthesia systems—including gas/vapour analyzers—should be designed to display partial pressure and not concentration.” In this chapter, the term P _MAC1 (see Table 3.1 ) is used to express the partial pressure of a potent inhaled anesthetic agent at a concentration of 1 MAC. Thus, 1 MAC of isoflurane is equivalent to a P _MAC1 of 8.7 mm Hg (see Table 3.1 ).

Latent Heat of Vaporization

Vaporization requires energy to transform molecules from the liquid phase to the gas phase. This energy is called the latent heat of vaporization and is defined as the amount of heat (calories) required to convert a unit mass (grams) of liquid into vapor. For example, at 20°C the latent heat of vaporization of isoflurane is 41 cal/g. The heat of vaporization is inversely related to ambient temperature in such a way that at lower temperatures, more heat is required for vaporization. The heat required to vaporize an anesthetic agent is drawn from the remaining liquid agent and from the surroundings. As vapor is generated and heat energy is lost, the temperatures of the vaporizer and the liquid agent decrease. This causes the vapor pressure of the anesthetic to decrease and, if no compensatory mechanism is provided, will result in decreased output of vapor. The temperature compensation mechanisms used in vaporizers are described in a later section.

Specific Heat

Specific heat is the quantity of heat (calories) required to raise the temperature of a unit mass (grams) of a substance by one degree of temperature (1°C). Heat must be supplied to the liquid anesthetic in the vaporizer in order to maintain the liquid’s temperature during the evaporation process, when heat is being lost.

Specific heat is also important when it comes to vaporizer construction material. Temperature changes are more gradual for materials with a high specific heat than for those with a low specific heat, for the same amount of heat lost through vaporization. Thermal capacity, defined as the product of specific heat and mass, represents the quantity of heat stored in the vaporizer body.

Also of importance is the material that is used in the construction of the vaporizer. Choosing the proper material allows the vaporizer to conduct heat from the environment to the liquid anesthetic. This property, called thermal conductivity, is defined as the rate at which heat is transmitted through a substance. In order for the liquid anesthetic to remain at a relatively constant temperature, the traditional variable bypass vaporizer is constructed from materials that have a high specific heat and high thermal conductivity. In this respect, copper comes close to the ideal (hence the Copper Kettle vaporizer). More recently, bronze and stainless steel have been used in vaporizer construction.

Measured Flow

The SVPs of halothane, sevoflurane, and isoflurane at room temperature are 243 mm Hg, 160 mm Hg, and 241 mm Hg, respectively. Dividing the SVP by ambient pressure (760 mm Hg) gives the saturated vapor concentration as a percentage of one atmosphere. This is an application of Dalton’s law (discussed earlier). The saturated vapor concentrations of halothane, sevoflurane, and isoflurane are therefore 32%, 21%, and 31%, respectively. These concentrations are far in excess of those required clinically (see Fig. 3.2 ; Table 3.2 ): therefore, the vaporizer first creates a saturated vapor in equilibrium with the liquid agent, and second, the saturated vapor is diluted by a bypass gas flow. This results in clinically safe and useful concentrations flowing to the patient’s breathing circuit. Without this dilution of saturated vapor, the agent would be delivered in a lethal concentration to the anesthesia circuit.

In the (now obsolete) measured flow (i.e., non–concentration-calibrated) vaporizers such as the Copper Kettle (Foregger/Puritan Bennett) or Verni-Trol (Ohio Medical Products), a measured flow of oxygen is set on a dedicated oxygen flowmeter to enter the vaporizer, from which vapor emerges at its saturated vapor concentration ( Fig. 3.3 ). This flow is then diluted by an additional measured flow of gases (i.e., oxygen, nitrous oxide, air, etc.) from the main flowmeters on the anesthesia machine (see Fig. 3.3 ). With this type of arrangement, calculations are necessary to determine the anesthetic vapor concentration in the emerging gas mixture that flows to the patient’s breathing circuit.

Contemporary anesthesia vaporizers are concentration-calibrated and most are of the variable bypass design. In a variable bypass vaporizer (e.g., Tec series from GE, Dräger Vapor 19.1, 2000 and 3000 series from Dräger) the total fresh gas flow from the anesthesia machine flowmeters (or electronic flow control system) enters the vaporizer ( Fig. 3.4 ). The incoming gas flow is split between two pathways. A smaller flow enters the vaporizing chamber (or sump) of the vaporizer and emerges with the anesthetic agent at its saturated vapor concentration. A larger flow bypasses the vaporizing chamber and eventually mixes with the outflow from the vaporizing chamber to create the desired or ‘’dialed-in’’ concentration (see Fig. 3.4 ) that flows to the patient circuit.

With both types of vaporizing systems, there must be an efficient method to create a saturated vapor in the vaporizing chamber. This is achieved by having a large surface area for evaporation. Flow-over vaporizers (e.g., Dräger Vapor 2000, 3000 series and GE Tec series vaporizers) increase the surface area using wicks and baffles. In the (now obsolete) measured flow, bubble-through vaporizers (i.e., Copper Kettle and Verni-Trol), oxygen is bubbled through the liquid agent. In order to increase the surface area, tiny bubbles are created by passing the oxygen through a sintered bronze disc (in the Copper Kettle). This creates large areas of liquid-gas interface over which evaporation of the liquid agent can rapidly occur.

Measured Flow Vaporizers (Copper Kettle, Verni-Trol)

Although measured flow vaporizers are not mentioned in anesthesia machine standards published after 1988 and are considered to be obsolete, it is helpful to first review the function of one such vaporizer, the Copper Kettle. Suppose that one wishes to deliver 1% (vol/vol) isoflurane to the patient circuit at a total fresh-gas flow rate of 5 L/min ( Fig. 3.5 ). This requires the vaporizer to evolve 50 mL of isoflurane vapor per minute (1% × 5000 mL) to be diluted in a total volume of 5000 mL.

In the Copper Kettle, isoflurane represents 31% of the atmosphere, assuming a constant temperature of 20°C and a constant SVP of 238 mm Hg. If 50 mL of isoflurane vapor represents 31%, the carrier gas (oxygen) must represent the other 69% (100%−31%). Thus

$\frac{50}{31}=\frac{x}{69}$

$x=\frac{50}{31}\times 69=111\text{mL}$

Therefore, if 111 mL/min of oxygen is bubbled through liquid isoflurane in a Copper Kettle vaporizer, 161 mL/min of gas emerges, of which 50 mL is isoflurane vapor and 111 mL is the oxygen that flowed into the vaporizer. This vaporizer output of 161 mL/min must be diluted by an additional fresh gas flow of 4839 (i.e., 5000−161) mL/min to create an isoflurane mixture of exactly 1% (because 50 mL of isoflurane vapor diluted in a total volume of 5000 mL gives 1% isoflurane by volume).

Although this situation is highly unlikely to occur in contemporary practice (because of the obsolescence of measured flow vaporizers), if one had to use a measured flow system to deliver isoflurane, the anesthesia provider would likely set flows of 100 mL/min oxygen to the Copper Kettle and 5 L/min of fresh gas on the main flowmeters, which results in very slightly less than 1% isoflurane (actually 44.9/5044.9 = 0.89%). Multiples of either of the vaporizer oxygen flow and main gas flowmeter flows would be used to create other concentrations of isoflurane from the Copper Kettle. Thus a 200 mL/min oxygen flow to the Copper Kettle vaporizer and 5 L/min on the main flow meters would create approximately 1.8% isoflurane. It is important to realize that if there is oxygen flow only to the Copper Kettle vaporizer and no bypass gas flow is set on the main machine flowmeters, lethal concentrations (approaching 31%) of isoflurane would be delivered to the anesthesia circuit, albeit at low flow rates.

Halothane and isoflurane have similar SVPs at 20°C (see Table 3.2 ), therefore the gas flows to be set for halothane would be essentially the same as those to be set for isoflurane when a 1% concentration of isoflurane is to be produced from a Copper Kettle. A Copper Kettle arrangement on an older model anesthesia machine is shown in Fig. 3.6 .

Enflurane and sevoflurane have similar vapor pressures at 20°C (175 mm Hg and 160 mm Hg, respectively), therefore similar flow settings could be used to create approximately the same agent concentrations with a measured flow system. In the case of sevoflurane, the measured flow vaporizer would contain 21% sevoflurane vapor (160/760 = 21%) ( Fig. 3.7 ). The oxygen flow therefore represents the remaining 79% of the atmosphere in the Copper Kettle. If precisely 1% sevoflurane is required at a 5 L/min total rate of flow, 50 mL/min of sevoflurane vapor needs to be generated. If 50 mL represents 21% of the atmosphere in the vaporizer, the carrier gas flow required is 188 mL/min [(50/21) × 79].

Thus if 188 mL/min of oxygen are bubbled through liquid sevoflurane contained in a Copper Kettle vaporizer, 238 mL/min of gas will emerge, 50 mL/min of which is sevoflurane vapor. This must be diluted by a fresh gas flow of 4742 mL/min (5000−238) to achieve exactly 1% sevoflurane.

Alternatively, using the formula given previously,

$\frac{160}{760}=\frac{50}{50+y}$

$y=188\text{mL}/\min$

where y = oxygen flow to the Copper Kettle vaporizer.

Setting an oxygen flow of 200 mL/min to the vaporizer and 5 L/min on the main flowmeters would result in a sevoflurane concentration of 1.01% [(200/79 × 21)/5253.2].

In the preceding examples, it was necessary to calculate both the oxygen flow to the measured flow vaporizer and the total main gas flow needed to produce the desired output concentrations of vapor. This was not only inconvenient, but also predisposed to errors that might result in serious overdose or underdose of anesthetic. Because of the obvious potential for error with a measured flow vaporizing system, if one were forced to use such a system, the concurrent continuous use of an anesthetic agent analyzer with high- and low-concentration alarms would be necessary to ensure patient safety.

Variable Bypass

In the concentration-calibrated variable bypass design of vaporizer, the total flow of gas arriving from the anesthesia machine flowmeters is split between a bypass and the vaporizing chamber containing the anesthetic agent (see Fig. 3.4 ). The ratio of these two flows, the splitting ratio, depends on the anesthetic agent, temperature, and chosen vapor concentration set to be delivered to the patient circuit. The required splitting ratio may be achieved either (1) by a control valve located upstream of the vaporizing chamber (i.e., where fresh gas enters the vaporizer) that controls the gas flow entering the vaporizing chamber (somewhat analogous to the measured flow vaporizing systems); or (2) by a control valve located downstream of the vaporizing chamber that proportions the flow of saturated vapor leaving the vaporizing chamber to mix with the bypass flow. Some older models of vaporizer were designed to create the flow split upstream of the vaporizing chamber (e.g., Fig. 3.8 ). In all contemporary variable bypass vaporizers the flow split is achieved by a control downstream of the vaporizing chamber ( Fig. 3.9 ). For all practical purposes, provided that the vaporizer is used at the ambient pressure at which it was calibrated (e.g., ∼760 mm Hg; sea level), it should make no difference whether the flow split is achieved by a control upstream or downstream of the vaporizing chamber. If, however, a vaporizer calibrated for use at ∼760 mm Hg ambient pressure is used under hyperbaric or hypobaric conditions, then the upstream location of the control valve will have a greater effect on performance. See later section on use of vaporizers under hypo- or hyperbaric conditions.

Fig. 3.10 depicts a variable bypass sevoflurane vaporizer set to deliver 1% sevoflurane at 20°C and 1 atmosphere pressure. A gas flow of 2079 mL/min entering the vaporizer must be split so that 79 mL enters the vaporizing chamber (where it constitutes 79% of the atmosphere because sevoflurane vapor constitutes an obligatory 21%) and 2000 mL enters the bypass. Emerging from the vaporizing chamber will be 21 mL sevoflurane and 79 mL carrier gas. The 21 mL of sevoflurane vapor is diluted in 2100 mL (2000+79+21), producing 1% sevoflurane. This results in a splitting ratio of 25:1 (2000/79) between the flow entering the bypass and the flow entering the vaporizing chamber. A variable bypass vaporizer (e.g., Dräger Vapor 2000) set to deliver 1% sevoflurane is therefore effectively set to a splitting ratio of 25:1 (see Fig. 3.10 ) for the inflowing fresh gas. (The term effectively is used because the flow split calculated for the incoming gas is really the result of the concentration control system acting downstream of the vaporizing chamber where it actually proportions the flow of saturated vapor exiting the vaporizing chamber to mix with the bypass flow.)

An alternative approach is to consider the gas mixture exiting the vaporizing chamber. Consider, for example, 100 mL/min of gas exiting the vaporizing chamber. This 100 mL comprises 21 mL of sevoflurane vapor and 79 mL of carrier gas. To create 1% sevoflurane by volume, the 21 mL of sevoflurane vapor must be diluted in a total volume of 2100 mL (because 21/2100 = 1%). The vaporizer control system acting downstream of the vaporizing chamber is therefore creating a flow ratio of 2000:100, or 20:1 between the bypass flow and the flow exiting the vaporizing chamber.

To summarize, when a contemporary variable bypass vaporizer is set to deliver 1% sevoflurane at 20°C, the control valve located downstream of the vaporizing chamber is creating a flow proportion of 20:1 between the bypass flow and the flow exiting the vaporizing chamber, thereby effectively creating a flow split of 25:1 (2000/79) between the bypass flow and the inflow to the vaporizing chamber.

As another example, consider a concentration-calibrated, variable bypass, isoflurane vaporizer set to deliver 1% isoflurane ( Fig. 3.11 ). What splitting ratio for incoming gases does this vaporizer achieve? The SVP of isoflurane at 20°C is 238 mm Hg, therefore the concentration of isoflurane vapor in the vaporizing chamber is 31% (238/760). If carrier gas enters the vaporizing chamber (where it now constitutes 69% of the atmosphere by volume, the other 31% being isoflurane vapor) at a rate of 69 mL/min, isoflurane vapor emerges at 31 mL/min and must be diluted in 3100 mL/min of total gas flow to produce a 1% concentration (since 31/3100 = 1%). Thus if carrier gas enters the vaporizer from the machine flowmeters at 3069 mL/min and is split such that 3000 mL/min enters the bypass while 69 mL/min enters the vaporizing chamber, when the gas flows merge, 1% isoflurane is the result. The inflow -splitting ratio is therefore 44:1 (3000/69) (see Fig. 3.11 ).

The flow split of 44:1 calculated for the inflowing gas is really the result of the concentration control system acting downstream of the vaporizing chamber where it proportions the flow of gas exiting the vaporizing chamber that mixes with the bypass flow. Assume that 100 mL/min of gas exits the vaporizing chamber. This 100 mL comprises 31 mL of isoflurane vapor and 69 mL of carrier gas. To create 1% isoflurane by volume, 31 mL of isoflurane vapor must be diluted in a total of 3100 mL. The vaporizer control system has therefore created a flow ratio of 3000:100, or 30:1 between the bypass flow and the flow exiting the vaporizing chamber.

To summarize, when a contemporary variable bypass isoflurane vaporizer is set to deliver 1% isoflurane at 20°C, the control valve located downstream of the vaporizing chamber is creating a flow proportion of 30:1 between the bypass flow and the flow exiting the vaporizing chamber, thereby effectively creating a flow proportion of 44:1 (3000/69) between the bypass flow and the inflow to the vaporizing chamber.

Table 3.3A shows the inflowing gas splitting ratios for variable bypass vaporizers used at 20°C. An equation for the calculation of inflowing gas splitting ratios is provided in the Appendix to this chapter.

Table 3.3B shows the flow ratios between the bypass gas flow and the flow of saturated vapor exiting the vaporizing chamber.

The concentration-calibrated vaporizer is agent-specific and must be used only with the agent for which the device is designed and calibrated. In order to produce a 1% vapor concentration, an isoflurane vaporizer makes an inflowing gas flow split of 44:1, whereas a sevoflurane vaporizer makes a flow split of 25:1 (see Table 3.3A ). If an empty sevoflurane vaporizer set to deliver 1% were filled with isoflurane, the concentration of the isoflurane vapor emerging would be in excess of 1%. Understanding splitting ratios enables prediction of the concentration output of an empty agent-specific variable bypass vaporizer that has been erroneously filled with an agent for which it was not designed. The change in concentration output when one-agent specific vaporizer is filled with a different agent can be calculated as the concentration set on the vaporizer dial multiplied by the ratio of the exit ratios (which is also the ratio of the SVPs of the two agents). In the case of the sevoflurane vaporizer set to deliver 1% sevoflurane (approximately 0.5 MAC) but filled with isoflurane, the resulting isoflurane concentration will be [1 × (30/20)], or 1.5% isoflurane (approximately 1.3 MAC). This can be potentially dangerous as the vaporizer is delivering 2.6 times the anesthetic potency that the user intended (1.3 MAC/0.5 MAC).

In summary, if a variable bypass vaporizer that has been calibrated for agent A and set to deliver a concentration C% is filled with agent B, the concentration C’ delivered by that misfilled vaporizer is calculated as:

Efficiency and temperature compensation

Agent-specific concentration-calibrated vaporizers must be located in the fresh gas path between the flowmeter manifold outlet and the common gas outlet on the anesthesia workstation. The vaporizers must be capable of accepting a total gas flow of 15 L/min from the machine flowmeters and of delivering a predictable concentration of vapor. However, as the agent is vaporized and the temperature falls, SVP also falls. In the case of a measured flow vaporizer (e.g., Copper Kettle) or an uncompensated variable bypass vaporizer, this results in delivery of less anesthetic vapor to the patient circuit. For this reason, all vaporizing systems must be temperature-compensated, either manually (with a Copper Kettle) or, as in contemporary vaporizers, automatically.

Measured flow vaporizers (e.g., Copper Kettle, Verni-Trol) incorporate a thermometer that measures the temperature of the liquid agent in the vaporizing chamber ( Fig. 3.6B ). A higher temperature translates to a higher SVP in this chamber. Reference to the vapor pressure curves (see Fig. 3.2 ) enables a resetting of either oxygen flow to the vaporizer, or the bypass gas flow, or both, to ensure correct output at the prevailing temperature. Such an arrangement, while tedious, does ensure the most accurate and rapid temperature compensation. The original Dräger Vapor vaporizer ( Fig. 3.12 ) (to be distinguished from the more recent Vapor 2000 and 3000 models fitted to contemporary Dräger anesthesia workstations) is a variable bypass vaporizer that incorporates a thermometer and a grid of lines on the vaporizer control dial for temperature compensation, whereby the desired output concentration is matched to the temperature of the liquid agent. Turning the control dial changes the size of an orifice in the bypass flow.

Most of the contemporary variable bypass vaporizers (e.g., GE-Datex-Ohmeda Tec series, Dräger 2000 and 3000) achieve automatic temperature compensation via a temperature-sensitive valve in the bypass gas flow. When temperature increases, the valve in the bypass opens wider to create a higher splitting ratio. More gas flows through the bypass, and less gas enters the vaporizing chamber. A smaller volume of a higher concentration of vapor emerges from the vaporizing chamber. This vapor, when mixed with an increased bypass gas flow, maintains the vaporizer’s output at reasonable constancy when temperature changes are not extreme.

Temperature-sensitive valves have evolved in design among the different types of vaporizers. Some older vaporizers (e.g., Ohio Calibrated Vaporizer) had, in the vaporizing chamber, a gas-filled bellows linked to a valve in the bypass gas flow (see Fig. 3.8 ). As the temperature increases, the bellows expands, causing the valve to open wider. Contemporary GE Tec series vaporizers (see Fig. 3.9 ) use a bimetallic strip for temperature compensation. This strip is incorporated into a flap valve in the bypass gas flow. The valve is composed of two metals each having a different coefficient of expansion (change in length per unit length per unit change in temperature). Nickel and brass have been used in bimetallic strip valves, brass having a greater coefficient of expansion than nickel. As the temperature increases, one surface of the flap expands more than the other, causing the flap to bend in a manner that opens the valve orifice wider, increasing the bypass flow. The principle of differential expansion of metals is applied similarly in the Dräger Vapor vaporizers ( Fig. 3.13 ), where an expansion element increases bypass flow and reduces gas flow in the vaporizing chamber as temperature increases. When temperature decreases, the reverse occurs.

The vapor pressures of the volatile anesthetics vary as a function of temperature in a nonlinear manner (see Fig. 3.2 ). The result is that the vapor output concentration at any given vaporizer dial setting remains constant only within a certain range of temperatures. For example, the Dräger Vapor 2000 vaporizers are specified as accurate to +0.20 vols% or +20% of the concentration set when they are used within the temperature range of 15°C to 35°C at one atmosphere of pressure. The boiling point of the volatile anesthetic agent must never be reached in the current variable bypass vaporizers designed for halothane, enflurane, isoflurane, and sevoflurane; otherwise, the vapor output concentration would be impossible to control and could be lethal.

Filling and Misfilling of Vaporizers

Contemporary concentration-calibrated variable bypass anesthesia vaporizers are agent-specific. If an empty vaporizer designed for one agent is filled with an agent for which it was not intended, the vaporizer’s output likely will be erroneous. Because at room temperature the vaporizing characteristics of halothane and isoflurane (SVP of 243 and 238 mm Hg, respectively), and enflurane and sevoflurane (SVP of 175 and 160 mm Hg, respectively) are almost identical, this problem at present mainly applies when halothane or isoflurane are interchanged with enflurane or sevoflurane.

Previously (see also ref. 12) it was shown that if a variable bypass vaporizer that has been calibrated for agent A and set to deliver a concentration C% is filled with agent B, the concentration C’ delivered by that misfilled vaporizer is calculated as

Bruce and Linde reported on the outputs of erroneously filled vaporizers at 22°C ( Table 3.4 ). Erroneous filling affects the output concentration and consequently the potency output of the vaporizer. In their study, an enflurane vaporizer set to 2% (1.19 MAC) but filled with halothane, delivered 3.21% (4.01 MAC) halothane. This is 3.3 times the anticipated anesthetic potency output. At 20°C, the predicted concentration of halothane would be 2.8% (3.7 MAC).

To summarize, if a vaporizer specific for an agent with a low SVP (e.g., sevoflurane) is misfilled with an agent that has a high SVP (e.g., isoflurane), the output concentration of the agent will be greater than that set on the concentration dial.

Conversely, if a vaporizer specific for an agent with a high SVP is misfilled with an agent that has a low SVP, the output concentration of the agent will be less than that indicated on the concentration dial. ^,

One must also recognize that the potency (MAC equivalent) of the agent concentration has to be considered in a misfilling situation. A sevoflurane vaporizer set to deliver 2% sevoflurane (approximately 1 MAC: see Tables 3.1 and 3.2 ) misfilled with isoflurane would produce an isoflurane concentration of 3% (approximately 2.6 MAC) (see Tables 3.1 and 3.2 ).

Erroneous filling of vaporizers may be prevented if careful attention is paid to the specific agent and the vaporizer during filling. A number of agent-specific filling mechanisms, analogous to the pin-index safety system for medical gases, are used in modern vaporizers. Liquid anesthetic agents are packaged in bottles that have agent-specific and color-coded collars ( Fig. 3.14 ). One end of an agent-specific filling device fits the collar on the agent bottle, and the other end fits only the vaporizer designed for that liquid agent. Although well intentioned, these filling devices cannot totally prevent misfilling.

A number of different filling systems are available (e.g., Quik-Fil [Abbott Laboratories, Abbott Park, IL], Key-Fill [Harvard Apparatus, Holliston, MA]). The agent-specific filling device for desflurane (Saf-T-Fill) is of particular importance because it is essential that a nondesflurane vaporizer never be filled with desflurane (see Desflurane section).

Vaporization of Mixed Anesthetic Agents

When funnel-fill vaporizer filling systems ( Fig. 3.15 ) were in common use, a more likely scenario was that an agent-specific vaporizer, partially filled with the correct agent, was topped up with an incorrect agent. This situation is more complex. It is much more difficult to predict vaporizer output, and large errors in concentration of delivered vapor could occur. Korman and Ritchie reported that, when mixed, halothane, enflurane, and isoflurane do not react chemically but do influence the extent of each other’s ease of vaporization. Halothane facilitates the vaporization of both enflurane and isoflurane and is itself more likely to vaporize in the process. The clinical consequences depend on the potencies of each of the mixed agents and on the delivered vapor concentrations.

Bruce and Linde reported that if a halothane vaporizer 25% full is filled to 100% with isoflurane and set to deliver 1%, the halothane output is 0.41% (0.51 MAC) and the isoflurane output is 0.9% (0.78 MAC) ( Table 3.5 ). In this case, the output potency of 1.29 MAC is close to the anticipated 1.25 MAC (1% halothane). On the other hand, an enflurane vaporizer that is 25% full and set to deliver 2% (1.19 MAC) enflurane and is filled to 100% with halothane has an output of 2.43% (3.03 MAC) halothane and 0.96% (0.57 MAC) enflurane. This represents a total MAC of 3.60, or more than three times that intended.

Table 3.5

Vaporizer Output After Incorrectly Refilling from 25% Full to 100% Full

From Bruce DL, Linde HW: Vaporization of mixed anesthetic liquids. Anesthesiology 1984;60:342–346.

Vaporizer	Setting (%)	Refill Liquid	Vaporizer Output						Total MAC
			Halothane		Enflurane		Isoflurane
			%	MAC	%	MAC	%	MAC
Halothane	1.0	Enflurane	0.33	0.41	0.64	0.38	—	—	0.79
	1.0	Isoflurane	0.41	0.51	—	—	0.90	0.78	1.29
Enflurane	2.0	Halothane	2.43	3.03	0.96	0.57	—	—	3.60
Isoflurane	1.5	Halothane	1.28	1.60	—	—	0.57	0.50	2.10

MAC, Minimum alveolar concentration.

It is important to avoid erroneous filling of vaporizers; if an error is suspected, the vaporizer should be emptied, withdrawn from service, labeled as misfilled, and returned to the manufacturer for servicing. ^,