^ *^ Sir- ^„* /a &*- * ^ -v^ftr. v * ^ *' • °* *' 7 / V^ T V V'^'V >*/^~V' %--.t.- v o V c5°^ .1 J? . * o°"«. ">>. :. >. «*"*++ ^^J&. : f* '.SUSS* ?**+ ~*%KH ^ * %*^>° V^'V & ♦sua* >& c* »/\Va'. ^ a* •ci^*. ^* ^ + A a* .vsGf. >. a* . w*» •«,. .,* .•»*•. v ,** .via ^ .4" ».^v^.» ♦♦*% ^ X9> "•^.T*" A n* -•»'•• "*c *4 \ w 4* ^<>» « *ivfi^^L «* «4 4> ..-. -H^ >°^ V*^V^ %*" #! ^V* *V # ^^V* V^' *?° £*» •" ♦♦' % ^^ 4 c» # r> # ^ 4? ^ - 1 <\ 6 6 *^ to o AjSfcA *.**&:* 4.* c° •!- ♦ AT 7!*, ^ ,0 4 »-^* V A> V .•"-• 9' ^Ofc ^ v ^i^:* ^ Ajr .«•• s°^ -of ^. •* a0 Q V**-.- .v ft' iA >/.: v^ *°« & ^ •: • A>^ a^ »'^K* %> A c . 3 The control station 5 Time considerations • 7 Contaminant transport time , i t 8 Tube traveltime , t£ 8 System sequence time, t S eq * H Design examples 12 Design example I 12 Design example II 14 Pneumatic monitoring for submicrometer particles 17 Summary and discussion 19 Conclusions 21 Appendix. — List of symbols 22 ILLUSTRATIONS 1. Schematic showing appropriate locations for auxiliary components along the length of a sampling tube 5 2. Generalized schematic of a pneumatic monitoring system 6 3. Block, diagram showing the electronic functions required for a pneumatic monitoring system 6 4. Characteristic flows for three small vacuum pumps as a function of the pressure drop across the pump 9 5. Maximum recommended tube lengths as a function of tube inside diameter for use in pneumatic monitoring systems 10 6. Schematic showing the relative locations of the sampling tubes and control station for the belt entry discussed in Design Example I 15 7. Schematic showing the relative locations of the sampling tubes and control station for the three returns discussed in Design Example II... 18 8. Maximum tube lengths as a function of inside tube diameter for pneumatic monitoring systems designed to detect submicrometer particles 19 DESIGN CRITERIA FOR RAPID-RESPONSE PNEUMATIC MONITORING SYSTEMS By Charles D. Litton } ABSTRACT This Bureau of Mines report presents a discussion of the essential components of pneumatic monitoring systems and their associated func- tions. Design criteria are presented which can be used for the design and fabrication of pneumatic monitoring systems having total system re- sponse times on the order of 15 to 30 min. To illustrate the utility of these design criteria, two detailed design examples are presented. ■Supervisory physical scientist/ Pittsburgh Research Center, Bureau of Mines, Pittsburgh, Pa. INTRODUCTION Improving the degree of safety afforded underground mine personnel is a major goal of the Bureau of Mines' research program. This report discusses in detail a methodology for monitoring of mine air contaminants that has the potential to improve underground mine safety by pro- viding early warning of developing haz- ards. The data on which this report is based was acquired from both in-house re- search projects and contractual efforts originating from the Fires and Explosions Group of the Pittsburgh Research Center, Bureau of Mines. Many potential hazards in underground mines are preceded by, or result in, the formation of contaminants that are car- ried throughout the mine by the imposed ventilation. Continuous monitoring of the mine air for contaminants has the po- tential to provide early warning of asso- ciated hazards in a time sufficient to successfully initiate control measures and to ensure the safety of underground personnel. Two techniques exist for continuous monitoring of the mine air. The first technique (called the electronic method) consists of placing one or more contam- inant sensors (called sensor packages) at carefully chosen locations within an un- derground mine. These sensor packages are then hard-wired to a remote station that provides electrical power for the sensors and accepts electrical signals from the sensors. The remote station may also contain a multiplexing function by which sensor signals are transmitted via two-conductor communication lines to a central, master control center. Depend- ing upon the size of the mine and the number of sensor packages, this method can be relatively simple (for instance, one sensor package with a remote station that also serves as the master control center); or it can be complex (for in- stance, several sensor packages with sev- eral remote stations controlled by a master control center). For this tech- nique, every underground monitoring point has a sensor package containing one or more sensors. In the second technique (called the pneumatic method), sensor packages are replaced by tubes. For every underground monitoring location, a tube extends from that location to a central control sta- tion. At this control station, pumps continuously pull samples of mine air from the monitoring locations through the tubes and then exhaust the samples into one or more contaminant sensors. Sensor outputs can be recorded and displayed at this station, or the outputs can be mul- tiplexed to a master control center for recording and display. The pneumatic method has been used suc- cessfully for continuous monitoring ap- plications where the rate of development of a particular hazard is slow with re- spect to the overall response time of the system. Such applications would include the continuous monitoring for detection of spontaneous combustion and for assess- ing the status of underground sealed fire areas. 2 It has been argued, and quite erroneously, that imposed time delays due to tube transit times and sequencing times are unacceptable for applications where the rate of development of a par- ticular hazard can be quite rapid. 2 Burgess, D., and H. Hayden. A Carbon Monoxide Index Monitoring System in an Underground Coal Mine. Soc. Min. Eng. , AIME, Ann. Fall Meeting, Salt Lake City, Utah, September 1975, SME Preprint 75- F-350, 25 pp. Chamberlain, E. A. C. , and D. A. Hall. The Practical Early Detection of Spon- taneous Combustion. Colliery Guardian, London, May 1973, pp. 190-194. Dalverny, L. E. , Z. J. Fink, and J. P. Weinheimer. Continuous Gas Monitoring Using Tube Bundles at the Joanne Mine Fire. BuMines TPR 92, 1975, 12 pp. While it is true that these times can be significant, it is also true that they can be dealt with quantitatively so that their effects can be minimized. Further, by presenting design criteria, primarily with respect to these time constraints, it becomes possible to calculate, in advance, the overall time response of a pneumatic monitoring system. And this information can subsequently be used to determine if a pneumatic monitoring sys- tem can be designed to meet the mon- itoring requirements for a proposed application. It is the intent of this report to describe the components necessary for fabrication of a pneumatic monitoring system and to present, in detail, design criteria that can be used to assess the limiting time responses of such systems. In so doing, it becomes apparent that, through prudent engineering design, pneu- matic monitoring systems can be used to monitor for hazards that develop quite rapidly. GENERAL DESCRIPTION A pneumatic monitoring system can be considered to be composed of two major elements. First, there is a set of two or more tubes extending from a central control station to the monitoring loca- tions. These tubes are used to convey air samples from the monitoring locations to the control station where various con- taminant levels are determined. Second, there is the central control station where pumps continuously purge the sam- pling tubes and present the air samples to the contaminant sensors for measure- ment. These two elements, and their com- ponents, will now be discussed in some detail. THE TUBING SYSTEM The set of tubes used for conveying air samples is commonly in the form of a bundle of tubes (hence, the name "tube bundles"), although for rapid-response systems, single tubing of >0.635-cm ID(> 1/4-in-ID) may often be required. Commercially available tube bundles con- sist of two or more tubes per bundle with each tube within a bundle having either an 0.427-cm ID(~l/6-in-ID) or an 0.635-cm ID (1/4-in-ID). The number of tubes per bundle ranges from 2 to 37 for the smaller size tubing, and from 2 to 19 for the larger size. Individual tubes are usually of polyethylene, although metal tubing may also be available from some manufacturers. Each bundle is cov- ered by an outer sheath of polyethylene (~l/8- to 1/4-in thick), which serves to protect the enclosed tubing. Bundles are generally in the form of reels, resem- bling reels of multiconductor electrical cable. Single tubing is also usually made of polyethylene; other tubing mate- rials are generally available, but at somewhat higher cost. Any single tube forms a path of com- munication between the monitoring lo- cation and the control station. Air samples must flow through a tube un- altered so that a contaminant measurement is truly indicative of the actual lev- els at the monitoring location. This constraint implies that some consider- ation should be given to the potential reactivity of contaminants during tube transport, either with the tube or with other gases in the air sample. In gen- eral, contaminants of interest, such as methane, CO, C0 2 , or smoke particles, are nonreactive and can be safely transported through polyethylene tubing. On the other hand, sulfur-containing contaminants, such as H 2 S or S0 2 , tend to react with either water vapor or condensed water during transport. Oxides of nitrogen can react with the polyethylene; if these gases are to be monitored, tubing made of non- reactive materials would be recom- mended. The more stable gases, nitrogen and oxygen, are nonreactive and can be safely transported through polyethylene tubing. In addition to the tubing, certain other components are generally required or recommended. At the sampling loca- tion, end-of-line dust filters should usually be attached to the sampling tubes to prevent accumulation of dusts within the tubes; this can sometimes result in clogging of the sample tubes. These filters are generally of large surface area and provide little resistance to the flow. If a tube bundle is used containing four or more tubes, it is generally rec- ommended that connections between tube bundles be protected by a junction box. A junction box is simply a rectangular box, made of heavy-duty metal or plastic, that can be mounted on the ribs of an entry. The lid of the junction box is easily removed, and tube bundles, one coming into the box from either side, can be connected and the lid replaced. The tube bundles entering the junction box are held firmly in place by brackets so that no tension is exerted on the con- nectors themselves. For bundles contain- ing fewer than three tubes, or for single tubing, it is recommended that, where connections are to be made, the tubing be firmly secured to the rib or roof on either side of the connection so that no tension is exerted on the connection. Once the connection is made, it should be wrapped with heavy-duty tape. Depending upon the location of the mon- itoring points relative to the control station, temperature differences along the path of a sampling tube may be en- countered; this can result in condensa- tion of water within the tubes. While water condensation within a tube generally will not affect the performance of the system, the transport of water into the contaminant sensors can signif- icantly affect their performance. For this reason, water traps are usually in- serted in the sampling line at the con- trol station just prior to the tube con- nection to the solenoid valves (use described below). In some applications, the gas sample may be flammable, and while no energy is available along the tube length to ignite such a flammable mixture, ignition could occur at the control station where power is supplied to the system. If ignition were to occur at this point, it is con- ceivable that flame propagation back through a sample tube could occur; this potentially could ignite a flammable gas mixture in the vicinity of the monitoring location. To avoid this potential prob- lem, simple in-line flame arrestors should be inserted in the sampling lines between the water traps and solenoid valves. It should be noted that no such occurrences have ever been observed, and these devices are recommended primarily as a precautionary measure. For applications where pneumatic moni- toring systems are to be designed for rapid response (see following sections), it is generally required that all sam- pling tubes have the same volumetric flow rate through them. For this reason, such systems should have manual valves in- serted in the sampling lines between the flame arrestor and the solenoid valves. These valves can then be used for obtain- ing the proper flow rate through each tube in the system. Such valves can be in the form of "needle" valves or "ball" valves; in most cases, the latter will suffice. If a sampling tube were fitted with all of these auxiliary components, then from the monitoring location up to the three-way solenoid valves, these com- ponents should be installed as per figure 1. Note that these components, except for dust filter and junction box, can be installed at the control station. THE CONTROL STATION The control station is literally the heart of a pneumatic monitoring system. It is at this station that pumps continu- ously purge all of the sampling tubes , with each tube sequentially connected in a regular, repeating fashion to the con- taminant sensors. A generalized, pneu- matic monitoring system is depicted in figure 2. Sampling tubes entering the control station are connected to the input ports of three-way solenoid valves (or their equivalent). Multiport rotary valves are available that can accomodate 12 to 24 incoming tubes. For systems using 12 or more sampling tubes, these valves can be used in the place of individual solenoid valves and at a reduced cost. One output port of each solenoid valve is connected to a large scavenger pump, while the second output port is connected to a smaller, sample pump. Typically, the scavenger pump continuously purges all of the sampling tubes, except for one, which is connected through the valving to the sampling pump. Each solenoid valve is energized in a regular, repeating fashion by a combination timer-controller (see figure 3) so that the contaminant levels within each tube can be determined. End of line dust filter Junction box r Control station Water Flame Ball tra P arrester VQ lve 3- way solenoid valve F | 0wt0 I scavenger pump Flow to sample pump FIGURE 1. - Schematic showing appropriate locations for auxiliary components along the length of a sampling tube. Not all components are required for each system. Sample pump Tubes from monitoring locations .Solenoid valves or equivalent ^Sensor Ij- (sensor 2j- ( Sensor nj >- Exhaust Scavenger pump Exhaust FIGURE 2. - Generalized schematic of a pneu- matic monitoring systemshowing the essential air sampling components at the control station. Once a solenoid valve is energized, the flow from that sampling tube is diverted to the sample pump, which in turn pro- vides flow to the contaminant sensors, as shown in figure 2. The number and type of sensors required will depend upon the requirements of the monitoring system and the information it is intended to provide. It is worth noting that, for this type of system, many contaminants can be measured at a single convenient location if so desired. Again, care should be taken in order to avoid any degradation of the air sample by components within the system. Pumps with nonreactive internal components are usually available and are generally rec- ommended for this type of system. Simi- lar consideration should be given to any component that is in direct contact with the air sample. In order for the control station to operate properly, it must contain elec- tronic functions. These functions are depicted in block form in figure 3. The combination timer-controller is used to provide power to the individual solen- oids. In general, the timer is set so that a single tube is sampled for a cer- tain period of time. At the end of that time interval, the timer generates a Timer controller Solenoid valves or equivalent Sample tube identifier (Alarm ^N outputs/^ Signal processor _jf Data "\ v^recorder/ Sensor outputs Contaminant sensors FIGURE 3. = Block diagram showing the elec= tronic functions required for a pneumatic moni = toring system. pulse that is received by the controller. The controller in turn produces a relay closure for the next solenoid valve, and so on. The timer-controller unit also produces a second output, which indicates the current tube whose sample is being sent to the sensors. This output, along with the outputs from the sensors, is connected to a signal processor. The signal processor in turn provides for alarm outputs and also for input to some type of data recorder, such as a strip-chart recorder. If remote alarms and remote data recording are required, the signal processor can also contain multiplexing and telemetry functions. This unit can be very simple (such as providing for alarms when contaminant levels exceed some value) or it can be more complex (such as providing for data recording and multiplexing or telemetry functions). In general, the signal pro- cessor should indicate the tube number associated with the alarm or data output. If desired, the signal processor can be used to provide information on the opera- tional status of the system. It can con- tain electronics that, for instance, supervise the sensors and other elec- tronic functions and provide indication of system malfunction. Tube integrity can also be monitored by inserting an electronic flowmeter in the line connecting the solenoid valves to the sample pump. This device can in turn be connected to the signal processor. Since all sampling tubes are required to have the same approximate flow, a low-flow indicator would signal that the tube is being blocked while a high-flow indi- cator would signal that a tube has been broken. Again, the number of electronic compon- ents and their complexity will depend upon the system requirements. This section has discussed the opera- tion of a generalized pneumatic monitor- ing system and the components required to make such a system operational. In prin- ciple, this information is sufficient to fabricate a pneumatic monitoring system. However, in order to design a system with a known response time, it is important that the various parameters that deter- mine the overall system response time be discussed. TIME CONSIDERATIONS A continuous monitoring system is in- tended to provide certain specific in- formation that can be used to signal the development of a potential hazard. For a monitoring system to function properly, it must be capable of providing this information in a time sufficient to suc- cessfully initiate control measures and to ensure the safety of underground personnel. Consequently, a monitoring system designed to protect against some hazard must have a maximum response time that is less than the anticipated devel- opment time of that hazard. If the hazard development time is x m and the maximum system response time is s » then T s < T m (1) if the system is to achieve its intended purpose. When some contaminant is re- leased into the mine air, that contami- nant is carried downstream at the venti- lation velocity, v f . If the monitoring location is some distance, £, from the point of origin of the contaminant, then the time necessary for the contaminant to B reach the monitoring location, if » is given by x t ■ 4/v f (2) Now, for a pneumatic monitoring system, once the contaminant reaches the monitor- ing location, it must travel through a tube of some length, l Q , before it reach- es the contaminant sensor. Provided that the pumps have sufficient capacity, the traveltime, t^ , for laminar flow (Rey- nolds number < 1,800) through a tube of length, l Q , in meters, and inside diameter, d Q , in centimeters, is given by 0.35 £ d . (3) Since all sampling tubes within the sys- tem should have the same volumetric flow rate, the maximum tube traveltime will occur for the longest sampling tube in the system. Once the contaminant reaches the cen- tral station, the maximum amount of time that it will take before the contaminant is detected will be one complete sequenc- ing time, t SE q. This results from the fact that this particular tube may have been sampled just prior to the contam- inant's reaching the central station, and should this happen, then the remaining tubes will be sampled before the contaminant-bearing tube again. If the sample time T SAMP> tnen tne total time through the tubes is is sampled per tube is to sequence and this time must satisfy equation 1 when used to monitor for a hazard that has a development time, x m . These three times will now be considered in detail. CONTAMINANT TRANSPORT TIME, t t The transport time, t + , (eq. 2) of a contaminant from its point of origin to a sampling point located a distance, £, downstream depends upon the value of I and the ventilation velocity, Vf , within the entry. Clearly, if i is zero, then x + is zero and this time need not be con- sidered. (See Design Example II.) How- ever, in many applications, the point of origin of the contaminant may not be known precisely, and sampling locations may have to be distributed in some logi- cal fashion in order to protect an entry (or entries) from a potential hazard. If, for instance, an entry is to be pro- tected from a hazard (such as fire) , and the point of origin of the associated contaminants could be any point along the entry, the sampling points should be spaced at some regular interval along the entry. If the entry length is £ E , then for n tubes, the spacing between sampling tubes, i D> will be given by *D = * E / n , (6) and the maximum transport time would be one spacing divided by the ventilation velocity, or SEQ = n t SAMP (4) Tf - *E nvf (7) where n is the number of tubes in the system. The sum of these three times, (eqs. 2- 4) is the maximum response time of the pneumatic monitoring system; that is, x s = A/v f + 0.35 Jl d + n r s AMP (5) TUBE TRAVELTIME, Tjl The tube traveltime, tjj, given by equation 3, depends not only upon the tube length and diameter but also upon the capacity of the pumps used to provide flow through # the tube. The volumetric flow rate, Q v , necessary to provide a traveltime, t^, is the tube volume di- vided by the tube traveltime; that is, 1,000 • lOOird* „ Qv = 7T- 2 - *o (8) with # £ in m, d Q in cm, t^ in seconds, and Q v in cm 3 /s. Substituting t^ from equation 3 yields Qv = 2 " d (9) Q corresponds to the volumetric flow for which the Reynolds number is ~1,800. Now, for a tube of length, l , side diameter, d Q , the required drop across that tube necessary vide a traveltime, t^ , can be be3 I 2 1 AP = P A -P+ = 0.076 ^7 — A * d o T l to pro- shown to (10) where P A = ambient and atmospheric pres- sure at the open end of the tube, assumed equal to 760 mm Hg (1 atm) P + = pressure at the end of the tube just before entering the pumps (mm Hg) The capacity of the pumps used must be capable of providing the flow, Q v (eq. 9), at the pressure drop, AP (eq. 10). In general, the small vacuum pumps used in these systems have free air capacities ranging from 236 cm 3 /s (0.5 ft 3 /min) to 7.1 x 10 3 cm 3 /s (15 ft 3 / min) , and are capable of contin- uous operation at pressure drops across the pump of 580 to 680 mm Hg. Such pumps show a linear decrease in flow rate with pressure drop, AP, ^Hertzberg, M., and C. D. Litton. Mul- tipoint Detection of Products of Combus- tion With Tube Bundles. Transit Times, Transmissions of Submicrometer Particu- lates, and General Applicability. Bu- Mines RI 8171, 1976, 40 pp. 800 600- E u S* o .cf 400 *- 200- 200 400 AP, mm Hg 600 FIGURE 4. = Characteristic flows for three small vacuum pumps as a function of the pres- sure drop across the pump. across the pump. Some typical pump curves are shown in figure 4, illus- trating this linear dependence on AP. These curves can be represented by the general expression Qv= Qo l- AP AP (ID Where Q = free air capacity (cm 3 /s) of the pump, and AP m = maximum pressure drop the pump can provide (mm Hg). Assuming the exhaust of the pump to be at atmos- pheric pressure, the pressure drop across the pump equals the pressure drop across the tube. By substituting the appropri- ate expressions for Q v , AP, and t : l> the required pump characteristics, Q and 10 AP, can be determined. The resulting expression is 225 d Q 1 - 0.22 i. (12) AP r V Equation 12 represents an explicit statement of the pump characteristics, Q (free air capacity), and AP m (maximum pressure drop), necessary for a tube of length i Qi in meters, and inside diam- eter, d Q , in centimeters, such that the traveltime, t^, can be expressed by equa- tion 3. Equation 12 should be used for determining the flow using the maximum tube length of a system, and is appli- cable only to flow through a single tube. Consequently, equation 12 should be used for determining the sample pump required for the system. Now, the scavenger pump must continu- ously purge all tubes except for one. Since it is required that the volumetric flow rate through all tubes be the same, then the capacity of the scavenger pump must satisfy Qscav > (n-1) Q (13) where n is the total number of tubes in the system. Equations 12 and 13 are important de- sign criteria for selection of pumps that are to be used in pneumatic monitoring systems for which the time response is crucial to the overall performance of the system. And, when pumps are used which satisfy these requirements, the tube traveltime of equation 3 can be used directly to determine the x^-component of the overall system response time. Equation 12 also provides a convenient means for determining the maximum length of a given tubing of inside diameter, d Q . As the denominator of equation 12 ap- proaches zero, the required pump capac- ity, Q , approaches infinity. By set- ting the denominator equal to zero, the maximum tube length as a function of in- side tube diameter can be determined as * MAX = 4.55 AP m d 3 (14) In practice, it is recommended that the maximum tube length not exceed 90% of this value in order to minimize pump requirements. Then, the maximum recom- mended tube length is U MAX >REC < 4.1 AP m d Q 3 (15) Figure 5 is a plot of (& 'REC versus d Q for an assumed AP m of 580 mm Hg. Using tube lengths in excess of their recommended lengths does not mean that flow through the tubes will cease. Flow will still occur, but at a much-reduced rate, resulting in longer tube travel- times. Further, when using these longer 0.5 1.0 1.5 INSIDE TUBE DIAMETER, d , cm FIGURE 5. - Maximum recommended tube lengths asa function of tube inside diameter for use in pneumatic monitoring systems. 11 lengths, it becomes difficult to deter- mine, in advance, what the values of these traveltimes will be, and equation 3 is no longer valid for these cases. In general, then, tube lengths greater than their recommended lengths should be used primarily in applications where the time response of the system is not crucial to its performance. Such applications might include the monitoring for spontaneous combustion, or the monitoring of sealed areas within a mine. For this type of application, the tube traveltimes can be measured and the time response of the system can usually be determined, even though it may not be crucial to the over- all intended purpose of the system. inside tube diameter is known, then this purge time is T M = *y*m x 6 4Q V (16) where the superscript "M" refers to the main exhaust line of the sample pump. Once the sample of gas reaches the "TEE" connector, it flows to the contam- inant sensor at a reduced rate, Q s , de- termined by the flow requirements of the sensor. This sensor purge time is given by ■ S - Trd s 2 JL 4Qs (17) SYSTEM SEQUENCE TIME, t seq For a pneumatic monitoring system com- posed of n tubes, the time to sequence from one tube through the remaining tubes and back to the original tube is given by equation 4. This time is determined by the number of tubes, n, and the sampling time per tube, tsamp* ^he sampling time per tube can be determined from a knowl- edge of the response times of the contam- inant sensors, and the time required to purge the tubing that connects the sample pump to the contaminant sensors. Generally, contaminant sensors are con- nected via a "TEE" connection to the ex- haust line of the sample pump (see fig- ure 2), with each sensor requiring a flow of 1 to 2 L/min. As a general rule, the purge time between samples from different tubes should be the time necessary to displace approximately 6 times the volume of the connecting tubing. The total flow rate through the exhaust of the # sample pump will equal the flow rate, Q v , pre- viously discussed. If the length of tub- ing, £ m , in centimeters, from the sample pump to the last "TEE" connector to the contaminant sensors is known, and the where the superscript "s" refers to the contaminant sensor tubing between the "TEE" connector and the sensor. In general, the sensor purge time is much greater than the purge time for the main sample exhaust so that in calculat- ing purge times equation 17 need only be considered. This purge time can be re- duced by using small lengths of tubing to connect the contaminant sensors to the sample pump exhaust, and by using smaller diameter tubing. Common 0.318-cm-OD (1/4-in-OD) tubing with a 0.43-cm-ID is recommended. Also, if the maximum response time, tr, of the contaminant sensors is known, then the interconnecting tubing should be of a length such that the additional purge time is less than about 20% to 25% of this response time. For instance, if the value of tr is 60 sec and the required sensor flow is 16.7 cm 3 /s (1.0 L/min), then for the above size tubing (0.43- cm-ID) , the maximum length of connecting tubing corresponding to 20% of 60 s is calculated from equation 17 to be 230 cm (~7.5 ft). Since, in general, sensors are located very close to the sample 12 pump, this condition can usually be sat- isfied with little effort. Then the total sample time is equal to the purge time plus the sensor response time; that is, T SAMP " T p + T R" (18) But, since x| must be less than 25% of t r , then the total sample time can be written in terms of response time: the maximum sensor T SAMP - 1>25 T R (19) Equation 19 is generally valid for sensors with response times greater than 30 s. If faster sensors are used, then equations 17 and 18 should be used to determine the sample time per tube. DESIGN EXAMPLES The preceding sections have discussed the components necessary for fabricating a pneumatic monitoring system and pre- sented guidelines for designing systems with rapid response times. This informa- tion can now be used for designing and fabricating pneumatic monitoring systems. In order to illustrate how this informa- tion can be used, the following two- exam- ples are given. These examples are in- tended to demonstrate the manner in which pneumatic monitoring systems can be de- signed to monitor for hazards that devel- op within a short period of time (~15 to 30 min) . Clearly, there exist many other potential hazards which are slower to develop and for which the pneumatic moni- toring approach is a viable and cost- effective technique. DESIGN EXAMPLE I It is desired to use the intake air from a conveyor belt haulageway to pro- vide additional ventilation at the work- ing coal face. In order to use the belt air for face ventilation, a carbon mon- oxide (CO) monitoring system with an alarm threshold of 10 ppm above ambient is required along the belt haulageway. The entry cross section is ~9.8 m 2 (7- by 15-ft) and is 1,800 m in length (~5,900 ft). The average ventilation velocity within the entry is 0.76 m/s (150 ft/min). Further, the minimum ac- ceptable alarm time for the CO monitoring system is 15 min (900 s). The objective, then, is to design a pneumatic CO monitoring system with a maximum response time of 15 min for a belt haulageway 1,800 m long with a ven- tilation flow of 0.76 m/s. To begin, a CO sensor must be chosen that is capable of responding at the 10- ppm-CO level and with a well-defined time response. Such a sensor is identified with a response time, t d , of 30 s. Then from equation 19, the tube is R» sample time per T SAMP = 37.5 s and the total sequence time, from equa- tion 4, will be T SEQ = 37 ' 5 n where the number of sampling tubes, n, is yet to be determined. The second step is to decide upon the location for the control station. One possibility is to locate the control sta- tion at the outby side of the belt entry. If this were to be done then the longest sampling tube would be approximately the length of the entry. From figure 5, it can be seen that the required tube di- ameter is ~0.9 cm (0.35 in) for this max- imum length of tubing. If these values for l Q (1,800 m) and d (0.9 cm) are in- serted into equation 3, then it is found that the tube travel time would be 567 s, 13 which leaves a total time of 333 s re- maining for contaminant transport and sequencing time; that is LL- + 37. 5n < 333. nv f If this equation is solved, using the values for £ E and v f , then it can be shown that no combination of n tubes can satisfy this time requirement. That is, any value of n will yield a total time greater than 333 s. Consequently, the control station must be located at some other point along the belt entry. For convenience, assume that the con- trol station can be located at the mid- point of the entry. Then the maximum tube length in the system will be 1/2 £ E or 900 m. From figure 5, the required tube diameter for this length is ~0.72 cm (0.28 in). This tube diameter is greater than 0.635 cm (1/4 in) but less than 0.953 cm (3/8 in). The closest standard size tubing that can be used is 0.794 cm (5/16 in). Substituting these values into equation 3 yields a maximum tube travel time of 250 s, which leaves a total time of 650 s remaining for contam- inant transport and tube sequencing; that is, "lL_ + 37. 5n < 650 Tube Travel Time *E nv 4 Substituting the appropriate values for £ E and v f , and rearranging yields 37.5n2-650n + 2,368 = Solving this equation for n indicates that when 5.2 < n < 12.1, the contaminant travel time plus the tube sequencing time will be less than 650 s. Consequently, the minimum number of sampling tubes which can be used is 6. Setting n=6, the following time components of the system result: Contaminant Transport Time Tp = 0.35 ^£_ d n = 250 s * nv f ° Tube Sequence Time T SEQ = 37. 5n = 225 s The maximum system response time is T S = T + + T^ + T SE Q = 870 S which is less than the required response time of 15 min (900 sec). Now, in order to select pumps to sat- isfy the tube travel time constraint, equation 12 indicates that, for a pump with a maximum AP m of 580 mm Hg (23 in Hg), its free air capacity must be greater than or equal to 562 cra 3 /s (~1.2 ft 3 /min); and equation 13 indicates that the free air capacity of the scav- enger pump must be > 2.81 x 10 3 cm 3 /s (~6.0 ft 3 /min). Since, in general, pumps with these capacities are readily avail- able, no problems are anticipated with regard to selection of pumps for this system. It should be noted that locating the control station at the midpoint of the haulageway does not represent the optimum location. The optimum location is that location which is central with respect to the number of sampling locations. In this case, the maximum tube length is given by W - 1/2 (*E - *£) - TJ *E and the total system response becomes ^£- + 0.35 ( 5^ ^ JUd n + 37. 5n. nvf y 2n J t ° Tc = t+ = A£- = 395 s nv Now, to satisfy the time response re- quirement, n must be >3. That is, if n is <3, then the contaminant transport time plus the sequencing time exceeds 900 s. Further, for n>4, the maximum tube length will be between 675 m and 14 900 m, and the closest standard size tubing that can be used for these lengths is 0.794 cm (5/16 in). For n=4, the sys- tem response time (from the above equa- tion) would be 930 s, which is too slow. However, when n=5, the system response time becomes 861 s, which is less than the required 900 s. Now, when n=5, the maximum tube length for a central control station is 720 m, and from equation 12, assuming AP m =580 mm Hg, the sample pump capacity required is 393 cm 3 /s (—0.84 ft 3 /min); and for the scavenger pump, 1.57 x 10 3 cm 3 /s (-3.4 ft 3 /min). Consequently, by locating the control station centrally with respect to the sampling tube locations, the number of tubes needed is five, rather than the necessary six if the control station were located at the midpoint of the belt entry. Also, this location would result in lower pump capacity requirements. While it is clear that the six-tube sys- tem would satisfy the time constraints for this application, the centrally lo- cated, five-tube system represents the optimum configuration. With the optimum five-tube system, there will be two tube lengths equal to 720 m and 0.794-cm-ID tubing will be used. There will also be two tube lengths of 360 m each. Referring to fig- ure 5, it can be seen that for these two sampling tubes, 0.635-cm (1/4-in) ID tub- ing will suffice. The fifth sampling tube is located at, or very near, the control station, so that even smaller tubing could be used if desired. For this situation, the pneumatic CO monitoring system can be defined as follows: station. Flame arrestors in each sample line are also recommended. 4. A local alarm shall be provided at the control station and provision made for a second remote alarm at the belt drive or other appropriate location. 5. The sample pump used will require a capacity of -393 cm 3 /s (0.84 ft 3 /min); and the scavenger pump, a capacity of 1.57 x 10 3 cm 3 /s (3.4 ft 3 /min). 6. Five three-way solenoid valves with associated sequencing controls are required. 7. A CO sensor with an alarm threshold of 10 ppm CO above ambient and a 30-sec response time will be used. 8. If desired, data can be acquired on a continuous basis via a strip-chart recorder or some other type of data acquisition system. A conceptual layout of the sampling lo- cations and control station for this belt entry is shown in figure 6. Before concluding this example, it is worth noting that an electronic system could also be designed for this applica- tion. Such a system would require three individual CO sensors spaced at intervals of 600 m (1,970 ft), with each sensor hard-wired to a remote control station. Assuming each sensor to have a time re- sponse of 30 s , then the maximum response time for this system is the contaminant transport time plus the sensor response time, or 820 s. This system's response time is -40 s less than the response time of the pneumatic system. 1. Five sampling tubes located at 360-m intervals along the belt entry, and connected to 2. A control station located 1,080 m (-3,540 ft) inby the belt drive. 3. Each sampling tube will require an end-of-line dust filter at the sampling location and a water trap at the control DESIGN EXAMPLE II Return airways from three working coal faces meet at a common point and air flows outside via a single, common re- turn. Return airway 1 (Al) is 1,600 m (-1.0 mi) long with a ventilation veloc- ity of 0.64 m/s (125 ft/min); the second return (A2) is 2,600 m (~1.6 mi) long with a ventilation velocity of 1.02 m/s 15 r < f < S / / 1,800 m |- 1.080 m i i— — l 360 m A y , y s y y y y y / / s y / / / / / / / / / / / s s ^S/ /<<{<< /V/SSSSf/SS 7 s S s y ' y Monitoring locations \ )>))))> j ; > >> ) ; > > } > } > 1 1 ) / ??>>>> f >> > r-n "ft Roof Air flow Conveyor belt 7 i t > n ? > > f ? s >> r /n />>>>> > }? t r >>> t »>>>) ) tnttftuwuttt Control station FIGURE 6. - Schematic showing the relative locations of the sampling tubes and control sta= tion for the belt entry discussed in Design Example I. (200 ft/min); and the third return (A3) is 2,500 m (~1.55 mi) long with a venti- lation velocity of 0.89 m/s (175 ft/ min) . The length of the common return is 1,600 m (~1.0 mi) with a ventilation velocity of 2.55 m/s (500 ft/min). Under normal conditions, the methane content in each of the returns is less than 0.1%. However, the coal seam is beginning to dip, and it is feared that higher methane concentrations may be encountered; this would necessitate ven- tilation changes and could substantially increase the potential for methane-air explosions. Further, each return con- sists of extensive abandoned, mined-out areas, and there is legitimate concern about the tendency of spontaneous fires to develop within these areas. A continuous monitoring system is de- sired that is capable of providing an alarm should the methane content in a return exceed 0.5%, or should a spontane- ous fire occur. The system must be capa- ble of continuous monitoring for methane (CH 4 ) in the 0% to 1% range, with an alarm at methane levels >0.5%, and must also be capable of monitoring for methane in the 0% to 15% range once alarm has been given. Also, the maximum response time of the system at the 0.5% CH 4 level should be less than 30 min, so that ade- quate time is available to correct any problems that may arise. It has also been decided that a CO sensor with an alarm threshold of 15 ppm CO above ambient and a response time of 30 sec is to be used for the detection of any spontaneous heatings that may occur. However, due to the long development times anticipated for the spontaneous heatings, the system response time could be much greater than 30 min and still be acceptable. Consequently, since both CO and CH 4 will be measured simultaneously, the system must be constrained to meet the 30-min alarm time for methane. The objective, then, is to design a pneumatic monitoring system that can sat- isfy all the above constraints. Since a CO sensor has already been identified, then an appropriate methane sensor must be selected. Such a sensor is identified with dual ranges of 0% to 1% and 0% to 15%, and a maximum response time of 40 s. Also, the methane sensor automatically switches to the 0% to 15% range should an alarm at the 0.5% level occur. Now, since both CO and CH 4 are to be measured, six sampling tubes will be required. Three sampling tubes will be located just inby the three working faces and are to be used primarily for methane measurements, although CO will also be measured simultaneously. Three other sampling tubes will be located just inby the intersection of the three returns and will be used primarily for CO measure- ments, although methane will also be mea- sured simultaneously. Because the meth- ane sensor has a longer response time, the sampling time per tube will be lim- ited by its response. From equation 19, 16 the sampling time per tube will be 1.25 x 40 s, or 50 s , and the total sequence time for six tubes (from equation 4) is 6 x 50 s or 300 s; this is one component of the overall system response time. An intake entry parallels the common return, and it is decided to locate the control station within this entry, such that the distance from the common inter- section to the control station is 50 m (~165 ft). The Al return is the shortest return, and it can be seen from figure 5 that tubing with a diameter greater than 0.93 cm (0.37 in) would be required for this length (~1,650 m). The standard size tubing that meets this requirement is 0.953 cm (3/8 in) ID. For either entry, A2 or A3, the tubing size must be greater than about 1.03 cm (0.41 in). The standard size tubing that meets this requirement is 1.11 cm (7/16 in) ID. For the three primary CO monitoring tubes, their lengths are approximately 60 m (~197 ft), so that these three sampling tubes must have inside diameters >0.26 cm (>0.103 in). It is decided that standard tubing with a 0.635-cm ID (1/4 in ID) will be used for these three tubes. A determination of the location for the three primary methane sampling tubes must now be made. Since the tube sequencing time is fixed at 300 s, then the total time available for methane transport from the face to the sampling tube, plus the travel time through a tube is 1,500 s (25 min) . For any one of the three en- tries, the total methane transport time is equal to the distance of the sampling point from the face divided by the venti- lation velocity within that respective* entry. Also, the maximum tube length for an entry is equal to the total entry length plus the 50 m to the control sta- tion minus the distance from the face. If the distance from a face to the sam- pling tube location is denoted by £, in meters, then the respective times for methane transport and tube travel for each entry (eq. 2 + eq. 3) become For Al, t A] - 1.563 l + 0.334 (1,650-*) For A2, x A2 = 0.98 i + 0.389 (2,650-jO For A3, x A3 = 1.124 l + 0.389 (2,550-Jt) When i = 0, the sampling tubes are lo- cated essentially at the face; from the above equations, it can be seen that this situation corresponds to the lowest times. The reason for this is that the samples are traveling through the tube at a velocity much greater than the entry ventilation velocity, so that the result- ant travel time is much lower. When the distance of a sampling tube from the face begins to increase, the resultant times begin to increase. Since the maximum time available is 1,500 s, the maximum distances from the face in each entry can be determined by setting the respective times equal to 1,500 s and solving for I for each entry. The results are I (Al) < 772 m I (A2) < 794 m I (A3) < 691 m So long as the distances of the sam- pling tubes from the faces are less than the above respective values, the total response time of the system will be less than the required 30 min (1,800 s). It is decided to locate each sampling tube a distance of 50 m inby the face of each of the three returns. Then the maximum sys- tem response times, to methane, within each return are t s (Al) = 913 s (15.2 min) x s (A2) = 1,360 s (22.7 min) t s (A3) = 1,330 s (22.2 min) and the lengths of tubing required for each entry are £ (Al) = 1,600 m i (A2) = 2,600 m 17 i (A3) = 2,500 m The pump requirements for the system will be determined by the flow require- ments for the longest tube (2,600 m). From equation 12, assuming a AP m equal to 580 mm Hg, the sample pump must have a capacity >896 cm 3 /s (>1.9 ft 3 /min). From equation 13, for this six-tube system, the capacity of the scavenger pump must be >4.48 x 10 3 cm 3 /s (>9.5 ft 3 /min). Since pumps with these capacities are generally available, no problems are an- ticipated with respect to pump selection. Then the pneumatic CH 4 -C0 monitoring system for this application is defined as follows: 1 . Two sampling tubes will be placed in both returns A2 and A3 (four total) with one sampling tube of 1.11-cm (7/16- in) ID located 50 m (-164 ft) inby the face, and the second sampling tube of 0.653-cm (1/4-in) ID located 10 m (~33 ft) inby the common intersection. 2. In entry Al , one sampling tube of 0.953-cm (3/8-in) ID will be located 50 m (~165 ft) inby the face and one sampling tube of 0.635-cm (1/4-in) ID will be lo- cated 10 m (~33 ft) inby the common intersection. 3. All six sampling tubes will be con- nected to the control station located in an intake entry approximately 50 m (~164 ft) from the common intersection. 7. A methane sensor with an alarm threshold of 0.5% and a response time of 40 s will be used to measure the methane levels from each return. The methane sensor will automatically switch to a 0%- 15% range should the 0.5% CH 4 level be reached. 8. Local alarms shall be provided at the control station with provision made for identification of whether the alarm is for CO or CH 4 and also identification of the sampling location producing the alarm. Provision shall also be made for a second, remote alarm with the same identifying feature at another appropri- ate location. 9. The sample pump used will require a capacity of ~896 cm 3 /s (1.9 ft 3 /min); and the scavenger pump, a capacity of 4.48 x 10 3 cm 3 /s (9.5 ft 3 /min). 10. If desired, provision can be made for continuous recording of CO and CH 4 data, either at the control station or at the remote alarm location. A layout of the planned system is shown in figure 7. These two examples clearly indicate that pneumatic monitoring systems can be designed to warn of hazards that develop fairly rapidly. PNEUMATIC MONITORING FOR SUBMICROMETER PARTICLES 4. Each sampling tube will require an end-of-line dust filter at the sampling location and both a water trap and flame arrestor at the control station in the sampling line just before entering that tube's three-way solenoid valve. 5. Six three-way solenoid valves with associated sequencing controls are required. Submicrometer ("smoke") particles are one contaminant that can provide the ear- liest indication of developing fires. 4 However, these particles, when trans- ported through tubes of a finite diam- eter, diffuse to the walls and hence can be lost completely within a sampling tube if certain precautions concerning tube diameter and tube travel time are not taken. For this reason, pneumatic 6. A CO sensor with an alarm threshold of 15 ppm CO above ambient and a response time of 30 s will be used to measure the CO levels from each return. ^Hertzberg, M. , C. D. Litton, and R. Garloff. Studies of Incipient Combustion and Its Detection. BuMines RI 8206, 1977, 19 pp. IS Intake entry LEGEND 2, 4 and 6 CH 4 sampling point /. 3 and 5 CO sampling point Common return Intake entry Return A3, 2,500 m Vx=0.89m/s Control station FIGURE 7. - Schematic showing the relative locations of the sampling tubes and control sta- tion for the three returns discussed in Design Example II. monitoring systems designed to measure the concentrations of submicrometer par- ticles have more stringent requirements on tube diameters and tube traveltimes. These additional requirements are pre- sented separately in this section. For submicrometer particles, it has been found 5 that sufficient constraint for pneumatic sampling is that 25% of all particles with diameters equal to 0.015 um be transmitted through the long- est sampling tube. For laminar flow in a tube (Reynolds number < 1,800), the fol- lowing relationship must be satisfied in order that this constraint be met:^ In order to satisfy both this con- straint and the laminar travel time con- straint of equation 3, the following re- lationship between the length and the tube diameter must be met: £ n SMP < 640 d, 'O (22) where the superscript "SMP" denotes the maximum tube length for pneumatic sam- pling of submicrometer particles. If equation 22 is compared to the maxi- mum recommended tube length (eq. 15), then it can be seen that the ratio is equal to 4D d ^ < 0.30 (20)* *a SMP 156 (Q™) REC AP m d (23) m"o For 0.015-u.m particles, the diffusion coefficient, Dj , has a value of 3.3 x 10 -l+ cm 2 /s 7 , so that the tube travel- time t£ , must satisfy 225 d/ (21) 5 Work cited in footnote 4. ^Fuchs, N. A. The Mechanics of Aero- sols. Pergamon Press Ltd./ London, 1964, pp. 184, 204-205. ^Work cited in footnote 6. Consequently, when d < 0.519 cm, as- suming AP m =580 mm Hg, the maximum tube length is limited by the pump capacity and when d Q > 0.519 cm, the maximum tube length is limited by the required trans- mission of smoke particles. In general, most rapid pneumatic monitoring systems will require standard tubing with >0.635- cm (1/4-in) ID; thus, most particle moni- toring systems will have their maximum tube lengths limited by equation 22. For convenience, the maximum tube length as a 19 function of inside tube diameter is plotted in figure 8, for pneumatic moni- toring systems designed specifically for measuring concentrations. submicrometer particle SUMMARY AND DISCUSSION In general, there exist two rather broad applications for monitoring sys- tems. One application (category 1) is for hazard detection along an entry (or entries) for which the point of origin of the hazard may not be well defined (for example, see Design Example I). In this type of application, the spacing of sampling points along the entry and the resultant contaminant transport time between sampling points contribute significantly to the overall response time of the system. For a pneumatic monitoring system with the central control station located at the mid- point of the entry, the response times can be estimated from the following equations: 2.! AetsampN 1/2 + 0.0105 l^'l for contaminant gases; or (24) > 2.1 f &ETSAMP 1/2 + 1.44 x 10 _1 + £ E 2 for submicrometer particles. (25) For instance, from Design Example I (£ E = 1,800 m, tsamp = 37.5 s, and Vf = 0.76 m/s), equation 24 would have predicted a system response time of ~856 s. The final calculated system response time for this example was 861 s. The second type of application (cat- egory 2) is for hazard detection when the point of origin of the hazard can be reasonably well defined. In this type of application, monitoring points are located in close proximity to the prob- able hazard origin so that contaminant transport times can be assumed negligible (for example, see II). The approximate a pneumatic monitoring instance, is given by «t/3 Design Example response time for system, in this T S > 0.030 AMAX + nTSAMP (26) for contaminant gases; or T S > 6.0 x 10"^ £§|AX + nTSAMP (27) for submicrometer particles. For instance, in Design Example II, airway A2 Umax = 2,650 m, n=6, tsamp = 40 s), equation 26 predicts a response time of 1,340 s, while the actual calcu- lated value was 1,360 s for the final design. These four equations can be used in making initial estimates of pneumatic 0.5 1.0 1.5 INSIDE TUBE DIAMETER,^, cm 2.0 FIGURE 8. - Maximum tube lengths as a function of inside tube diameter for pneumatic monitoring sys- tems designed to detect submicrometer particles. 20 monitoring system response times to de- termine if this type of system can be used in a particular application. In general, if the estimated value of i s for a proposed application is within ±10% of the required response time, x m , or if the estimated value of t s is significantly less than t m , then a pneumatic monitoring system should be more seriously consid- ered for the proposed application. In formulating a decision regarding the potential design and fabrication of a pneumatic monitoring system, the following initial questions should be addressed: 1. What is the intended function of the monitoring system? 2. What is the maximum hazard response time, x m , that can be tolerated in this application? 3. To which applications category does this application belong? Category 1 — Wide-area coverage for which the point of origin of the hazard is not well defined; or Category 2 — Localized coverage for which the hazard origin is reasonably well defined. 4. If the application belongs in cate- gory 1 , then for each entry to be moni- tored, the entry length and average entry ventilation velocity should be specified and used in either equation 24 or equa- tion 25, depending upon whether the con- taminant to be monitored is a gas or sub- micrometer particles, respectively. In utilizing either equation 24 or 25, an average value of t$amp = 45 s will gener- ally suffice. In some instances, an entry may be too long so that a single central control station with associated monitoring points may not provide sufficient time response. In these instances, some thought should be given to subdividing the entry, with each subdivision having its central con- trol station and associated monitoring points. The example, if a proposed application requires that x m < 1,200 s (20 min) and the entry is defined by % £ = 4,830 m (3 miles) and v f = 1.02 m/s (200 f t/min) , then x s (from eq. 24), would have an estimated value of ~1,830 s (30.5 min), which is considerably longer than the required response time of 1,200 s. How- ever, if the entry is subdivided into two equal lengths (2,415 m each), then two central control stations would be re- quired with each subsystem having an estimated time response of ~1,026 s (17.1 min) , which is much less than the re- quired response time of 1,200 s. Once x s nas been verified for a pro- posed category 1 application, then the number of monitoring points required for this application can be estimated from (x m -0.012 iE */3) - y then it is reasonable to assume that a pneumatic monitoring system can be successfully implemented for this application. CONCLUSIONS The basic components of pneumatic moni- toring systems have been described and detailed design criteria presented for rapidly responding pneumatic monitoring systems. The design examples show how this information can be used in the design of such systems. The information contained in this report should be suffi- cient for designing pneumatic monitoring systems for many potential applications. 22 APPENDIX. —LIST OF SYMBOLS Dj subraicrometer particle diffusion coefficient, cm 2 /s. d m inside diameter of connecting tubing from sample pump to "TEE" connector, centimeters. d Q inside diameter of a pneumatic sampling tube, centimeters. d s inside diameter of tubing connecting the contaminant sensor to the main pump exhaust line, centimeters. I distance from point of origin of a contaminant to a downstream pneumatic sampling point, meters. Iq the distance between two consecutive pneumatic sampling parts, meters. £ E the length of a mine entry, meters. l m the length of tubing connecting the sample pump exhaust to the "TEE" connector, centimeters. £ MAX the maximum sampling tube length within the pneumatic sampling system, meters. l the length of a pneumatic sampling tube, meters. £, MAX the maximum sampling tube length, in meters, which can be used with a fixed sampling tube inside diameter, d Q , and a pump with a maximum rated pressure drop, AP m , in mm Hg. (£ max )rec tne maximum recommended sampling tube length, in meters, for use with a fixed sampling tube inside diameter, d Q , and pump with a rated maximum '0 'KtL, _ w.^J k, pressure drop, AP m . U MAX )rec = 0.90 I. " £ SMP the maximum sampling tube length, in meters, which can be used for pneumatic sampling of submicrometer particles through a sampling tube of inside diameter, d Q . l s the length of tubing connecting the contaminant sensor to the main sample pump exhaust line, centimeters. n the number of sampling tubes for a pneumatic sampling system. n min the estimated minimum number of sampling tubes required to provide a system response time equal to, or less than, some required, maximum response time. P A atmospheric pressure = 760 mm Hg. P.J. pressure at the end of a sampling tube prior to entering the sample pump, mm Hg. 23 AP the difference between P A and P + , (P A -P t ), mm Hg. AP m the maximum rated pressure drop for a given pump, mm Hg. Q free air capacity of the sample pump, cm 3 /s. Q s flow rate required by a contaminant sensor, cm 3 /s. Qscav free air capacity of the scavenger, or purge, pump, cm 3 /s. Q v required volumetric flow rate, cm 3 /s, through a sampling tube of inside diameter, d Q , and length, i Q , Vf ventilation air velocity within a mine entry, m/s. To the maximum travel time, seconds, for laminar flow through a tube of inside diameter, d Q , and length, S, Q . T m the maximum anticipated hazard development time, seconds. Tp the time required to purge the tubing connecting the sample pump to the "TEE" connector, seconds. T p the time required to purge the tubing connecting the contaminant sensors to the main sample pump exhaust line, seconds. tr the time response of the contaminant sensor, seconds. t s the maximum calculated response time for a pneumatic monitoring system, seconds. T SAMP tne sampling time per individual sampling tube within the system, seconds. t$eq the time required to sequence through all of the system's sampling tubes, seconds. T+ the time required for the contaminant to travel in the ventilation flow from its point of origin to a pneumatic sampling point, seconds. INT.-BU.OF MINES, PGH., PA. 26555 >°v * ^ V \\/ .-itor.V** .-Mter-. V./ .•isM**. ^** .-^flter-.V^ •■- y %'.*%*? y v^V V*^--V %''^-/ V^'V.. %. '.ft 4V* 1 ,* V; W ". V . ■ • w '.•^t.'V /..-^>o /..^I-X ,/,C^>o >»\ v °o > " \* .. * •■* * -4* «A V ' i~ .r v.* J> o • O J"" < rP* .1^ ^°v «<> »i V ^-^K V ^^ ^ ..v^-*b o A *^r« # ^ ^ *-t^ .* p V - « • , ^_. MAY 83 I WS2F N - MANCHESTER, ' "•-* INDIANA 46962 —— — -^ 'iimmiBmiiiiiiiy Bliii 1BH HHHHHL IIIhIHI nllHillll ■'■■■.•;■■■■'■'::■■:.;. HBBHflBBgBgra HanHSSSMBHieBl&HfaniHS ■■BHP flwwWaSwBowBRWiaa JiS :■■:■■■■■.•:>;■. nnHr Wmmm&l WWH^ 889 g mm -.■■'=■ --: : iv- HI H ■n, |«n nliliililS ,ngr& sS