British Journal of Anaesthesia 1994; 72: 474-^479 The desflurane (Tec 6) vaporizer: design, design considerations and performance evaluation R. B. WEISKOPF, D. SAMPSON AND M. A. MOORE SUMMARY We have described the design and design considerations of the desflurane Tec 6 "vaporizer" and have tested its performance characteristics. The vaporizer differs from previous vaporizers designed for anaesthesia in that electromechanical rather than mechanical controls accommodate the different physical characteristics of desflurane. This design, while offering perhaps an increased risk of failure (owing to sophisticated electronic components and circuitry), on the other hand offers the decreased likelihood of accidental delivery of very large concentrations of liquid anaesthetic resulting from tilting or overfilling and alarms and warnings not previously incorporated into the design of anaesthetic vaporizers. The output characteristics of the vaporizer are as expected, based on the design: desflurane concentration output in oxygen has accuracy (±75%) which is similar to that of the mechanical vaporizers; output decreases when nitrous oxide is added owing to the lower viscosity, but remains within 20% of the dial setting or 0.5% absolute. (Br. J. Anaesth. 1994; 72: 474-479) KEY WORDS Anaesthetics, volatile: desflurane. Equipment: vaporizers. The new anaesthetic, desflurane (1,2,2,2-tetrafluorethyl-difluorethyl ether), has a relatively large vapour pressure of 669 mm Hg at 20 C and 757 mm Hg (boils) at 22.8 C. Current temperature-compensated, agent-specific vaporizers were not designed to accommodate anaesthetics with a boiling point or a very steep vapour pressure vs temperature curve within its expected temperature of operation (fig. 1) [1]. This report describes the design, considerations that went into the design of a new "vaporizer" created specifically to deliver desflurane and an evaluation of the accuracy of its output. DESIGN AND CONSIDERATIONS Compatibility It was thought advisable that a new vaporizer should have external design and performance characteristics similar to those of current vaporizers. The vaporizer would have to be able to be mounted, in a plug-on fashion, on the "back bar" of current types of anaesthesia machines. These considerations dictated vaporizer height and width, but allowed modest flexibility in depth. Also dictated was the incorporation of a "safety interlock", which prevents the simultaneous use of more than one vaporizer. Perhaps most important was the need to provide a dial indicating the concentration of desflurane output of the vaporizer. Concentration had to be affected little by changes in background flow, room temperature and background gas composition. The MAC of desflurane (7.25% in oxygen, in humans aged 18 30 yr) suggested the incorporation of a relatively large sump, while the very low blood-gas solubility (0.42) of desflurane indicated a need to be able to fill the vaporizer during the delivery of an anaesthetic, without having to shut the output of the vaporizer (to avoid substantial decrease of depth of anaesthesia). Size limitations suggested the latter approach. Temperature control The options are either to cool or to heat the liquid anaesthetic to some constant temperature to produce a constant vapour pressure. Failure to do so would result in large variations in temperature of the liquid anaesthetic. If the anaesthetic were cooled it would have been possible to use the same stream-splitting technology now used in the variable bypass vaporizers. However, this was not practical because few commercially available methods for cooling are compatible with the size constraints for the vaporizer, and those that are available either consume much power or are unreliable, or both. In addition, failure of cooling could result in high vaporizer temperatures, with the potential for excessive anaesthetic output, requiring additional potentially cumbersome systems to safeguard against this possibility. Consequently, the alternative approach was taken: to heat desflurane to a constant temperature of approximately 39 C, thus providing a constant vapour pressure of approximately 1460 mm Hg, and therefore delivery of desflurane as a gas. This pressure was chosen as sufficient to overcome the resistances of the internal components of the vaporizer (the "regulator", the rotary valve and passage RICHARD B. WEISKOPF, M.D. Department of Anesthesia, Cardiovascular Research Institute, University of California, San Francisco, CA 94143-0648, U.S.A. DAVE SAMPSON, B.SC., Research and Development, Ohmeda, Steeton, West Yorkshire, U.K. MARK A. MOORE, B.A., University of California, San Francisco, CA, U.S.A. Accepted for Publication: October 21, 1993. Correspondence to R.B.W.
DESFLURANE VAPORIZER 475 failure, the vaporizer ceases to deliver anaesthetic and the battery provides power exclusively for the warning and status lights for approximately 20 h (see below). The vaporizer draws a maximum current of 2.5 A. Thus if plugged into a multiple socket outlet available on some anaesthesia machines, total consumed current may exceed the combined socket rating if other electrical devices are in operation. 800-, Pb at sea level ex 700-2 ^ 6003 o FEATURES Q. 50015 16 17 20 18 19 Temperature ( C) 21 I 23 22 FIG. 1. Vapour pressure of desflurane at temperatures of 15-23 C. ways within the vaporizer; see fig. 2) and of the breathing system, and also provide an extra margin of safety in the event of partial occlusion of the breathing system. To provide the necessary heat, two elements heat the sump containing the anaesthetic and two more heat the vapour in the outflow path (see below) to prevent condensation. Liquid anaesthetic temperature is monitored by a resistive thermometer device (RTD). The electronics determine the amount of power to be supplied to the heaters. A second RTD monitors die temperature of the heater block within the sump and shuts off power to the heaters if the temperature exceeds 57 C. The vaporizer is powered by 110 V or 220 V a.c. It contains a 9-V battery which is of insufficient power to supply the electricity required for heating, which is variable, but is approximately 50 W during steadystate conditions and 200 W maximum to permit addition of room temperature anaesthetic, without affecting output. Thus in the event of main power The functional aspects of the vaporizer may be thought of as being composed of two portions: (1) components constituting a regulator function (shown in the shaded area of fig. 2); and (2) the other components handling the gases: inlet, fixed restrictor, sump, shut-off valve, manual variable control valve and outlet. The gases comprising the "background flow" follow a path through a fixed resistor, and unlike current Tec 3, 4, and 5 vaporizers, leave the vaporizer without ever having entered the sump containing the anaesthetic. The sump is heated (and, as a result, pressurized). Its output of gaseous desflurane is controlled by a series of transducers, valves and resistors. The "sump shut-off" valve opens when the vaporizer is enabled and the anaesthetist-controlled dial ("manually variable restrictor ") is rotated to any position other than " 0 % " desflurane (as detected by an optical sensor). The sump shut-off valve is closed by any of several conditions (see below), when cessation of delivery of desflurane is desired or mandated. The anaesthetic vapour then passes through the regulator (shown in shaded area of fig. 2) and then through the manually controlled variable restrictor, after which it joins the background gases and leaves the vaporizer. The manually controlled variable restrictor is controlled by the anaesthetist. It comprises a groove Fixed restrictor R1 Fixed restrictor R2 Carrier gas inlet Alarm control electronics Pressure Temperature level Battery Mains system Control electronics LJ I Differential pressure tri sducer - Output Jilt Dial I 1 Interlock L_J solenoid Liquid ll level X, Sump shut-off value Heated sump Variable I restrictor Regulator Manually variable restrictor select Temperature control F I G. 2. Schematic d r a w i n g of the vaporizer courtesy of O h m e d a, I n c..
476 BRITISH JOURNAL OF ANAESTHESIA cut on the underside of a plate of the external dial mounted on top of the vaporizer, on which the calibrated scale of concentrations of desflurane appears. Only desflurane vapour passes though this groove. As the dial is turned, the portion of the groove which is exposed to desflurane vapour has an increasingly deep cut, allowing for greater flow. THE REGULATOR The purpose of the regulator is to reduce the pressure of desflurane vapour (1460 mm Hg) leaving the sump (fig. 2, point D) and make the resulting pressure equal to the pressure of the background gas. This is accomplished by an electromechanical system, which was chosen rather than a simple mechanical device for several reasons: (1) the ease of achieving the range of vapour flow (between 2 and 2200 ml min" 1 ); (2) the linear pressure-flow characteristic; (3) size constraints; and (4) electronic control permitted warning, alarm ("operational", "no output", "warm-up" and "alarm battery low") and control functions not otherwise achievable. The differential pressure transducer senses pressure at point B in the background gas flow, upstream of the fixed restrictor and at point E in the desflurane vapour flow. When the pressures are not equal, an electrical signal is sent to the variable restrictor of the regulator to correct the discrepancy. If the pressure in the desflurane flow is less than that in the background flow, the valve opening is increased to allow for greaterflowof desflurane; conversely, if the pressure in the desflurane flow exceeds that of the background gas, the opening of the variable restrictor is decreased, to decrease the flow of desflurane. Because of the fixed, laminar flow restrictor (R2), an increase in background gas flow increases the pressure in the proximal portion of the background flow channel TABLE I. Viscosity of gases commonly used in anaesthesia (from [2]) 240 - E 220 - (O Viscc Gas Oxygen Nitrous oxide Nitrogen Carbon dioxide Helium 280 -, 260-200 - 180-160 - l_l or ^ 206 x 10-8 at 19.1 C 152 x 10" 8 at 26.9 C 182 x 10-8 at 27.4 C 153 x 10" 8 at 19 C 198 x 10" 8 at 20 C ^ Viscosity (kg s" 1 m" 2 ) 120 - i j i 1 ' 1 ' 1 ' 1 260 280 300 320 340 360 380 400 Temperature (K) FIG. 3. Viscosity of oxygen ( ) and nitrous oxide (O)- (point B), creating an imbalance of pressures between the background gas channel and the desflurane channel (point E). This is sensed by the differential pressure transducer, and an electrical signal causes the variable restrictor to open further, increasing the flow and pressure of desflurane in the channel. Conversely, decreasing the background gas flow results in an opposite imbalance at the pressure transducer and a resultant electrical signal which decreases theflowand pressure of desflurane at point E. For safety, there is a second differential pressure transducer (not shown in fig. 2) which senses pressures at the same locations. If the two transducers do not agree, an electrical signal is sent to the sump shut-off solenoid, which stops desflurane output from the sump. This redundancy in the system also safeguards against zero drift in the transducers. Should the variable restrictor (R2) in the desflurane channel be open fully, but the pressure at point E in the desflurane channel still be less than that in the background gas channel (point B), additional power is supplied to the sump heaters. Ordinarily this is required only when the sump is filled with additional anaesthetic while the vaporizer is delivering large desflurane concentrations with large flows of background gas. For the above to function satisfactorily, the fixed restrictor (R2) in the background gas channel and the rotary variable restrictor ("manually variable restrictor select", fig. 2) must have laminar flow so that any change in pressure produces a linear change in flow. Within the working limits specified (0.2-10 litre min" 1 ) the flow through these devices is essentially laminar. However, the flow through the fixed restrictor, as is true for all laminarflowdevices, is affected by the viscosity of the gas. The vaporizer is calibrated at the factory with a backgroundflowof 100% oxygen. At any pressure change across the resistor (R2), as the viscosity of the gas increases, the flow of gas decreases. For example, because nitrous oxide has a lower viscosity than oxygen (table I, fig. 3), as it is added to the background gas, the viscosity of the background gas decreases and thus for a given gas pressure at point B (proximal to the fixed restrictor, R2), background gasflowis greater than if the background gas flow were oxygen alone. As the regulator balances pressures (and not flows) in the two gas channels, at this pressure (but greaterflowin the background gas channel) the amount of desflurane allowed to flow though the rotatory variable control valve is unchanged, resulting in a decreased concentration of desflurane leaving the vaporizer. The change in pressure decrease and flow is directly proportional to the change in viscosity. Fortunately the viscosities of the other gases used commonly in anaesthesia are sufficiently similar to that of oxygen [2] (table I) that the resultant desflurane concentration emanating from the vaporizer when nitrous oxide or nitrogen is added, should remain within 20 % of the value when oxygen alone comprises the background flow (see below). The background gas channel contains one additional laminar flow, fixed resistor (Rl) in line before the gas enters the regulator component of the vaporizer. The purpose of this resistor is to provide
DESFLURANE VAPORIZER 477 a small positive pressure (19-22 cm H 2 O at an oxygenflowof 10 litre min"')j on the side of the plate within the anaesthetist-controlled dial (manually variable restrictor) of the vaporizer, in order to transmit the pressure at point A to the plate. Desflurane vapour passes through the groove on the opposite side of this plate. The positive pressure helps to prevent the plate from lifting. If there were no additional resistor, and if the plate were to lift, then the groove through which the desflurane passes would have a larger effective cross-sectional area, allowing a greater than desired amount of desflurane to pass. With this additional restrictor any lift of the rotary valve results in a net inflow of fresh gas into the control groove area, causing a dilution of desflurane output. SAFETY FEATURES Several safety features are incorporated into the Tec 6 desflurane vaporizer. A second differential pressure transducer (not shown in fig. 2) measures the difference in pressures at the same place as the first transducer. The two must read identically, or the shut-off solenoid is activated, and output of desflurane ceases. This guards against zero drift of both transducers. Both resistive thermometer devices (one for the sump, and one for the upper manifolds and rotary valve) have back-ups (not shown in fig. 2). As for all pressurized devices, a "relief" valve is desirable; with the Tec 6 this comprises the use of "thermal fuses", one in each limb of the main power supply to the sump heaters (not shown in fig. 2). If the temperature exceeds 57 C, the power to the heater is interrupted, thus preventing build up of excessive pressure. To ensure that the vaporizer does not deliver desflurane when the anaesthetist controlled dial is turned to 0% (stand-by), not only is the resistor (groove) in the dial closed, but in addition, a positive interlock mechanism shuts the sump shut-off valve which controls output from the sump. Before the control dial valve can be rotated to a position other than 0%, the dial release button must be depressed. When the dial release button is pushed in, an optical sensor sends a signal to open the solenoid controlling the sump output. An interlock prevents the release button from being depressed unless the vaporizer is in operational mode. A second button on the anaesthetist-controlled dial must be depressed to deliver desflurane concentrations in excess of 12%. If tilted, some conventional, mechanical vaporizers are subject to excessive output, owing to liquid anaesthetic entering the tubes (designed to carry vapour only) leaving the sump. Because the sump shut-off solenoid opens only when the vaporizer is operational and thus electrically powered, and placed correctly on the back bar of an anaesthesia machine, the likelihood of liquid desflurane leaving the sump when the vaporizer is tilted is reduced greatly. Nevertheless, the Tec 6 also contains a "tilt switch" which shuts the sump shut-off valve (should it be open) when the angular displacement of the vaporizer exceeds approximately 15 from vertical. The Tec 6 desflurane vaporizer incorporates electronic sensors and display of the hquid level of the sump. This method was chosen, rather than the direct visual " sight glass " of conventional vaporizers because of the possibility of potential for leaks caused by the heated pressurized liquid anaesthetic. The liquid level sensor is a ceramic covered electrode within a stainless steel tube, which is part of the sump. The capacitance of this sensor, which forms one arm of a comparator, varies with the level of liquid between the ceramic covered electrode and the stainless steel tubes. The electrical signal through the electronics drives the 20-bar liquid crystal linear display on the front of the vaporizer. When the sump is full, all 20 bars are illuminated; when the sump contains less than 60 ml of hquid, no bar is illuminated. A second device switches off the heaters when the sump is almost empty. This results in a no output alarm, which changes to a low agent alarm when the rotary dial is returned to a standby position. Overfilling the sump of a standard mechanical vaporizer may result in liquid anaesthetic entering the collecting tubing exiting the sump. In conventional vaporizers, because background gas flows through the sump, this liquid is carried out of the vaporizer rapidly, either as liquid or a very large concentration of vaporized anaesthetic. This has resulted in several reports of anaesthetic overdose. Because no background gas flows through the sump of the Tec 6 vaporizer, this risk is lessened. However, it is important to avoid overfilling of the sump. There are two mechanisms designed to prevent overfilling of the sump. First the filler aperture of the sump is placed so that when the level of desflurane in the sump is above the aperture, thus occluding it, the sump is full, but not overfilled. Second, the sump outlet is higher than the liquid level of a full bottle of desflurane in itsfillingposition. Thus under ordinary circumstance, the sump fluid level would not reach this point. However, it should be noted that the design of both of these mechanisms depends upon the vaporizer being filled in its normal upright position. Filling the vaporizer while it is in another attitude could bypass this design and result in overfilling. However, even if the vaporizer were overfilled while tilted, the solenoid shut-off valve would prevent liquid from leaving the sump. When the vaporizer was subsequently turned on, a small amount of liquid might leave the sump. However, the resultant vapour would cause an increase in pressure in the desflurane gas channel, resulting in closure of the shut-off solenoid. EVALUATION OF OUTPUT PERFORMANCE We have (1) compared the desflurane concentration in the gas leaving the vaporizer with the dial setting; (2) measured desflurane output as a function of background gas flow rates; and (3) measured desflurane output as a function of the background gas composition. Methods We tested six desflurane Tec 6 vaporizers at desflurane dial settings of 1, 4, 7 and 10% with
478 BRITISH JOURNAL OF ANAESTHESIA a 10 CO T3 8-5 2H Q. g g a 9 8 -t-4 i A i 4 S- S 5 T i a " i i x -a 9 r 2 4 6 8 Gas flow (litre min" 1 ) -9 fr A i i 10% 7%» 4% FIG. 4. Desflurane output of six vaporizers (each represented by a different symbol) at vaporizer settings of 1, 4, 7 and 10% desflurane with various carrier gas flow rates of 100% oxygen. c 5= O 8-6-.H 2 - Q. 5 0 10 8 8 I 8 S B 8 10% g i 7% i i n 8 4 % 2 4 6 8 Gas flow (litre min" 1 ) FIG. 5. Desflurane output of six vaporizers (each represented by a different symbol) at vaporizer settings of 1, 4, 7 and 10% desflurane with various carrier gasflowrates of 70 % nitrous oxide in oxygen. I 25-, -M I «* ft ft- 20-10 H 5- o «o 2 4 6 Gas flow (litre min" 1 ) FIG. 6. Difference (relative; %) in vaporizer output at dial settings of 1 (O), 4 ( )» 7 (O) and 10% (A) desflurane, between carrier gasflowsof 100 % oxygen and 70 % nitrous oxide in oxygen. Data are mean of six vaporizers. background flow rates of 0.5-10 litre min" 1 of 100 % oxygen and 70% nitrous oxide in oxygen. To determine desflurane concentration in the gas leaving the vaporizer, we used the analogue output, to the nearest 0.01 % desflurane, of a clinical infra-red spectrophotomer (Datex Ultima, Helsinki), which 10 1% 10 had been calibrated against gas samples, the concentration of which had been measured by gas chromatography against primary gravimetric standards. In addition, spot samples of vaporizer output were analysed by gas chromatography. Results With carrier flow of oxygen 0.5-10 litre min" 1, the desflurane concentration in the outflow gas was within ±15% relative or ±0.5% absolute, of the dial setting for all vaporizers, except one, which was 17 % low at the single condition of 4 % desflurane at 0.5 litre min" 1 (fig. 4). At carrier flow rates less than 2 litre min" 1 of oxygen, desflurane concentration was generally smaller (by approximately 8%) than at larger carrier gas flow rates, but was still within the above limits. With carrier gas flow rates of 70% nitrous oxide in oxygen, the concentration of desflurane was less (average 16% smaller) than with similar carrier flow rates of oxygen (figs 5 and 6). The desflurane concentration in the output gas was within these limits for all vaporizers at a dial setting of 1 % desflurane. At larger desflurane concentrations, and larger carrier flow rates of 70 % nitrous oxide in oxygen, the desflurane output of the vaporizers remained within these tolerances; however as the carrier gas flow rate decreased, so did the relative output of the vaporizers. Only at a dial setting of 1 % desflurane did output error exceed 20 % (range 0-32 %; mean error (SD) : 13.6 (6.5) %); however, output remained within ±0.5% desflurane, absolute. Discussion The accuracy of the Tec 6 vaporizer output is dependent upon the desflurane concentration, and carrier gas flow rate and composition. The vaporizer is calibrated at the factory with oxygen as the carrier gas. Thus it is not surprising that the vaporizer output closely matched the dial settings when oxygen was the carrier gas. Because of the lower viscosity of nitrous oxide and the vaporizer design, a smaller desflurane concentration is to be expected when nitrous oxide is part of the carrier gas than when 100 % oxygen is the carrier gas. At 20-22 C, the viscosity of nitrous oxide is 71 % that of oxygen [2]. Thus the viscosity of 70 % nitrous oxide in oxygen should be approximately 80% that of 100% oxygen. This should result in an approximate 20% decrease in the concentration of desflurane leaving the vaporizer when 70 % nitrous oxide in oxygen is the carrier. At a dial setting of 1 % desflurane, we found that the desflurane concentration with 70% nitrous oxide in oxygen as the carrier gas was 20% less than the desflurane concentration with 100% oxygen as the carrier gas. As the dial setting was increased the difference in desflurane concentration between the two carrier gases lessened to 10-15 % at 7-10 % desflurane. The small difference in desflurane output between analysed and theoretical values when nitrous oxide was added to the carrier gas, at any dial setting, suggests that although flow within the fresh gas restrictors is essentially laminar, it is not perfect. Fortunately, the error is in the direction of safety: as the second
DESFLURANE VAPORIZER 479 anaesthetic (nitrous oxide) is added, the vaporizer output is less than indicated by the dial setting. REFERENCES 1 Andrews JJ, Johnston RV jr, Kramer GO Consequences of A /~vwnurr irrw-cuxexrr misfilling contemporary vaporizers with desflurane. Canadian Journal of Anaesthesia 1993; 40: 71-76. The authors thank E. I. Eger n, M.D., for his review of this 2 Hodgman, Weast RC, Selby SM, eds. Handbook of manuscript. This work was supported by the Anesthesia Chemistry and Physics. 39th Edn. Cleveland: Chemical Research Foundation. Rubber Publishing Co, 1957; 2043-2047.