On-Line Measurement of Heat of Combustion of Gaseous Hydrocarbon Fuel Mixtures

Transcription

1 NASA Technical Paper 3572 On-Line Measureent of Heat of Cobustion of Gaseous Hydrocarbon Fuel Mixtures Danny R. Sprinkle Langley Research Center Hapton, Virginia Sushil K. Chaturvedi and Ali Kheireddine Old Doinion University Norfolk, Virginia National Aeronautics and Space Adinistration Langley Research Center Hapton, Virginia March 996

2 Available electronically at the following URL address: Printed copies available fro the following: NASA Center for AeroSpace Inforation National Technical Inforation Service (NTIS) 800 Elkridge Landing Road 5285 Port Royal Road Linthicu Heights, MD Springfield, VA (30) (703)

3 Abstract A ethod for the on-line easureent of the heat of cobustion of gaseous hydrocarbon fuel ixtures has been developed and tested. The ethod involves cobustion of a test gas with a easured quantity of air to achieve a preset concentration of oxygen in the cobustion products. This ethod involves using a controller which aintains the fuel (gas) voluetric flow rate at a level consistent with the desired oxygen concentration in the cobustion products. The heat of cobustion is deterined fro a known correlation with the fuel flow rate. An on-line coputer accesses the fuel flow data and displays the heat of cobustion easureent at desired tie intervals. This technique appears to be especially applicable for easuring heats of cobustion of hydrocarbon ixtures of unknown coposition such as natural gas. Introduction Several well-known techniques are available currently for the easureent of heat of cobustion of hydrocarbon fuels. Constant-volue bob calorieters are coonly used for deterining heat of cobustion of liquid fuels (refs. and 2). Siilarly, constantpressure flae calorieters are eployed for easuring the heat of cobustion of gaseous fuels (refs. 3 and 4). Singh et al. (ref. 5) have developed an on-line technique which requires injection of pure oxygen prior to cobustion for easuring the heat of cobustion of gaseous fuels. The injection is necessary to aintain the required 2-percent oxygen concentration in the cobustion products. Singh et al. have also developed a siplified ethod that eliinates the need for the addition of pure oxygen and uses a coputational algorith that is valid only for a prescribed air flow rate and cobustionproduct oxygen concentration (ref. 6). The present technique offers an iproveent over this latter ethod by using a coputational algorith that is valid over a wide range of flow rates and oxygen concentrations. Developent of the present easureent syste satisfies the need for accurate, on-line easureent of heat of cobustion of gaseous fuel ixtures with ethane as the doinant coponent. This fuel has been used in the cobustor of the Langley 8 Foot High-Teperature Tunnel (8-Ft HTT) to produce teperature levels approaching 2000 K. In the present study, an on-line gas calorieter was developed and tested for high-precision easureent of heat of cobustion of gaseous hydrocarbon ixtures. The ethod described herein involves the easureent of the concentration of oxygen in the cobustion products and its correlation with the heat of cobustion of the fuel. Sybols A = 4 f i x i + ( + P) f i y i 4P i = C f C x H y E f f i fuel flow rate correction factor hydrocarbon optical encoder data total hydrocarbon ole fraction ole fraction of ith hydrocarbon species I inert fraction of fuel l air voluetric flow rate nuber of hydrocarbon species n fuel voluetric flow rate P cobustion-product oxygen ole fraction P o ole fraction of oxygen in air R 2 coefficient of deterination α, β regression coefficients H heat of cobustion Experiental Setup The prototype gas calorieter is shown scheatically in figure and photographically in figure 2. The fuel gas is ixed with air and fed into a specially designed burner. The fuel flow rate is easured by a voluetric floweter with an optical encoder ounted on its shaft. The flow rate is deterined by easuring the counts/in of the etched lines on the encoder disk. Each revolution of the shaft, which is noinally equivalent to 000 cc, produces 200 counts. A burner that is fabricated with an internal honeycob structure was used to ensure a stable flae over a range of air and fuel flow rates. The burner was enclosed in a vented glass container which prevented atospheric containation of the cobustion products. The oxygen concentration in the cobustion products was easured by puping the cobustion products through a yttriu-stabilized zirconiu oxide oxygen sensor that produces a voltage related to the concentration of oxygen in the environent to which it is exposed (ref. 7). The oxygen sensor output was directed to a

4 controller that aintained the desired concentration of oxygen in the cobustion products by varying the fuel flow rate. This technique requires a sufficient quantity of oxygen (>0. percent) in the cobustion products to be easurable by the oxygen sensor. A set point concentration of 3-percent oxygen was used throughout this study. Because of the precision uncertainties of the voluetric floweter, the oxygen sensor, and the air floweter, and because of the inherent unsteadiness of the cobustion process, the oxygen concentration in the product gases exhibited sall fluctuations about the 3-percent set point. To account for the rando fluctuations in the oxygen concentration, the ean and the standard deviation (of heat of cobustion) were obtained for a set of data points at several tie intervals. Experiental Procedure The oxygen sensor was calibrated with ixtures of known oxygen concentrations so that a sensor output (in V) versus oxygen-concentration calibration curve could be obtained. Subsequently, the syste was flushed with air for several inutes to scavenge any fuel that was left inside the syste fro the previous experient. Air fro a pressurized cylinder was used for cobustion, and its oxygen concentration was easured with the oxygen sensor prior to the experient. The fuel floweter was also calibrated so that the nuber of counts/in fro the optical encoder could be converted readily into flow rate in standard cubic centieters per inute (scc). A given air flow rate fro the pressurized cylinder was initiated, and the fuel flow rate was gradually increased to a level at which a stable flae was established. The 3-percent oxygen concentration was dialed as a set point in the controller, and the controller was activated to aintain the fuel flow rate. The fuel flow-rate data in counts/in were recorded at regular intervals and were used in a correlation (described in the Test Results section) to obtain the value of heat of cobustion. Over a tie period, several data points were recorded, and the ean and the standard deviation were coputed for the period. Theoretical Principles The heat of cobustion H ix of a gaseous ixture that has several saturated hydrocarbons and a noncobustible inert gas ay be expressed as H ix = f f i H i () where f is the total hydrocarbon ole fraction, is the nuber of hydrocarbon species, f i is the ole fraction of the ith hydrocarbon species ( f i = ), and H i is the heat of cobustion of the ith hydrocarbon species. The cobustion of this ixture with air ay be expressed as l+ n[ f( f C x H y + f 2 C x2 H y2 + + f C x H y ) + ( f)i] Heat nf ( f x + f 2 x f x )CO 2 + nf ( f 2 y + f 2 y f y )H 2 O nf + lp o nf ( f x + + f x ) ( f 4 y + + f y ) O 2 + [ l( P o )]N 2 + n( f)i where l is the voluetric flow rate of air, n is the voluetric flow rate of the gaseous fuel ixture, I is the noncobustible inert fraction of the test gas, and P o is the ole fraction of oxygen in the air. The ole fraction of oxygen in the cobustion products P can be expressed as Solving for f, f (2) (3) (4) Substitution for f in equation () yields the following expression: Let P 4lP o 4nf f i x i nf f i y i = nf f i y i + 4n 4nf + 4l 4 lp ( o P) 4nP = n 4 f i x i + ( + P) f i y i 4P f i H i 4lP ( H o P) ix = P n 4 f i x i + ( + P) f i y i 4P (5) A is a easure of the effective hydrocarbon content of the cobustible coponent. Note that for saturated Air is assued to be coposed of nitrogen and oxygen. A = 4 f i x i + ( + P ) f i y i 4P i = 2

5 hydrocarbons, A can be further siplified since y i =2x i +2. The H ix ay now be expressed as 4lP ( o P) H ix = P n i = f i H i A (6) For a binary ixture of a single pure hydrocarbon species ( = ) and an inert gas, the above equation reduces to 4lP ( o P) H H ix = P (7) n A where A =4x+(+P)y 4P, and H is the heat of cobustion of the hydrocarbon species. Note that for ethane (y = 4), A is independent of P. For a ixture of unknown coposition, the value of f i H i /A to be used in equation (6) is not known since f i and A are unknown. Table I shows values of H (ref. 8), A, and H /A for various saturated hydrocarbons. (When ore than one for of a hydrocarbon exists, such as n-butane and isobutane, the average value of H is given.) A value of 0.03 for P is used in the calculation of A in each case. In the present technique, the adoption of a universal value of the f i H i /A coefficient in equation (6) is crucial for easuring the heat of cobustion of a gaseous ixture of unknown coposition. However, as noted fro table I, H/A varies fro for hexane to 26.6 for ethane. For the present technique it is sufficient to adopt a value of this coefficient that will iniize the error in the prediction of the H ix for the doain of saples to be tested. Thus, for the doain of saples where all species occur with equal probability, the ean value of is the appropriate choice for f i H i /A in equation (6). In nature, however, hydrocarbons are not distributed uniforly. Methane akes up the bulk of the hydrocarbon coponent in ost natural gas saples. Figure 3 Hydrocarbon Table I. Values of H, A, and H/A for Pure Saturated Hydrocarbons H, * kcal /ol A (P = 0.03) H/A Methane, CH Ethane, C 2 H Propane, C 3 H Butane, C 4 H Pentane, C 5 H Hexane, C 6 H Mean = ± 0.5 * All heats of cobustion H are given as the gross heat of cobustion (producing liquid water at 25 C (ref. 8)). All errors in this paper are given at the 95-percent level of confidence. shows the variation of f i H i with the ethane ole fraction of the hydrocarbon coponent for 265 well saples that were analyzed in a 982 Bureau of Mines survey of natural gases (ref. 9). Figure 4 shows the variation of f i H i /A for these saples. Although the value of f i H i varies significantly fro saple to saple, the value of f i H i /A varies in a narrower range. Figure 5 presents a frequency distribution plot of f i H i /A for these saples. More than 90 percent of these well saples have a f i H i /A between 26.5 and Thus, the ean value of is the appropriate choice for f i H i /A for a wide range of natural gas saples. Table II shows values of f i H i, A, and f i H i /A for six natural gas pipeline saples fro the Bureau of Mines survey. Although values of f i H i and A vary significantly fro saple to saple, the value of f i H i /A is constricted to a narrow band fro to 26.59, in agreeent with the ean value cited previously. Thus, although the adoption of a universal value for the f i H i /A coefficient inevitably introduces an error in the calculated value of H, this error will be negligible in the overwheling ajority of saples fro natural gas sources. Table II. The f i H i, A, and f i H i /A for Six Natural Gas Pipeline Saples Hydrocarbon coponent concentrations, * ole percent f i H i, A No. CH 4 C 2 H 6 C 3 H 8 C 4 H 0 C 5 H 2 C 6 H 4 kcal/ol (P = 0.03) f i H i /A Trace Mean = ± 0.04 * Fro reference 9. 3

6 Test Results In the present study, l = 7000 scc and P = For these values equation (6) reduces to 28000( P o 0.03) H ix = n Equation (8) fors the basis for calculation of the H ix fro the voluetric fuel flow rate easureents. The present technique was first applied to two pure gas saples: ethane and ethane. Since heats of cobustion of these gases were already known, the results were used to deterine the precision error of the ethod, and to this end a correction factor C f was introduced in the calculation of voluetric flow rate fro the optical encoder data E to correct for bias in the syste. Table III shows the experiental data for ethane that involves optical encoder data E in counts/3 in intervals, the fuel flow rate n in scc calculated fro the optical encoder data (n = E/3 in * 000 cc/200 * C f ), and the heat of cobustion ( H) in kcal/ol calculated fro equation (8); P o was easured at The ean value Table III. Methane Data i = f i H i A E, counts/ 3 in Fuel flow rate, n, * scc H, kcal/ol Mean = 68.3 ± 3. Mean = ±.22 * Cf =.053. (8) of the heat-of-cobustion data is ±.22 for ethane. Table IV shows the data for ethane; P o was easured at Readings were taken at 5-in intervals. The ean value of the heat-of-cobustion data for ethane is ± These two cases indicate that a precision approaching 0.6 percent can be achieved for pure saturated hydrocarbon fuels such as ethane and ethane. Table IV. Ethane Data E, counts/ 5 in Fuel flow rate, n, * scc H, kcal/ol Mean = ± 2.5 Mean = ± 2.60 * Cf =.007. The syste was also used to find the heat of cobustion of a ixture that was percent ethane and 7.5-percent nitrogen, which has a calculated heat of cobustion (eq. ()) of kcal/ol; P o was easured at Table V shows data for this ixture, along with the fuel ole fraction f calculated fro Table V. Data for Percent CH 4 Plus 7.5 Percent N 2 E, counts/ 3 in Fuel flow rate, n, * scc H, kcal/ol Fuel ole fraction, f (calculated fro eq. (4)) Mean = ± 0.3 Mean = ± 3.0 Mean = ± * Cf =

7 Table VI. Suary of Data for Methane, Ethane, and Methane Plus Nitrogen H est = E Regression output: α, ; Standard error of H est, ; R 2, ; Nuber of observations, 33; Degrees of freedo, 3; β, ; Standard error of β, Calibration gas Flow data (E), counts/3 in H, kcal/ol H est, kcal/ol Ethane Methane Methane, percent equation (4). A value of for H and a value of 8 for A were used since ethane was the only hydrocarbon involved. The ean value of the heat-of-cobustion data is ± 3.0. The ean value of the fuel ole fraction data is ± Nuerical Approach The heat of cobustion ay also be expressed as H = α + β Ē -- (9) where E is flow data, either in the for of flow units or directly in optical encoder counts, and α and β are constants deterined fro a least-squares regression analysis of flow data and heat of cobustion for gases of various heat content, assuing constant values for l, P o, and P. Equation (9) allows calibration of H directly in ters of optical encoder counts using calorietric gas standards, thus obviating the interediate steps of quantifying the flow rate and oxygen concentration of the cobustion air, calibrating fuel flow rate fro optical encoder counts, and calibrating oxygen concentration fro oxygen sensor voltage. This regression algorith 5

8 was applied to optical encoder readings for the ethane, ethane, and ethane-plus-nitrogen data presented above and effectively provides a 3-point calibration of H in ters of E (optical encoder data for ethane were noralized for 3-in intervals). Table VI suarizes these data and presents results of the regression analysis. A graphical plot of the data, the least-squares regression curve fit H est, and the ±2 standard error of H est (SEE) band, which approxiates the 95-percent confidence interval about the curve fit (ref. 0), are shown in figure 6. Conclusions An experiental technique has been developed and tested to easure the heat of cobustion of a saturated hydrocarbon and inert gas ixture. The technique is especially attractive for applications in which an on-line easureent of the heat of cobustion of natural gas is required. The syste worked reasonably well during the course of laboratory investigation and thus validated the technique. Preliinary test results indicate that this technique can deterine the heat of cobustion of a gas of unknown coposition with an uncertainty on the order of 2 percent. NASA Langley Research Center Hapton, VA January 22, 996 References. Anon.: Standard Test Method for Heat of Cobustion of Liquid Hydrocarbon Fuels by Bob Calorieter. Annual Book of ASTM Standards, ASTM, June 992, pp Anon.: Standard Test Method for Heat of Cobustion of Hydrocarbon Fuels by Bob Calorieter (High-Precision Method). Annual Book of ASTM Standards, Dec. 988, pp Anon.: Standard Test Method for Calorific (Heating) Value of Gases in Natural Gas Range by Continuous Recording Calorieter. Annual Book of ASTM Standards, Apr. 994, pp Anon.: Standard Test Method for Heating Value of Gases in Natural Gas Range by Stoichioetric Cobustion. Annual Book of ASTM Standards, Mar. 989, pp Singh, Jag J.; Sprinkle, Danny R.; and Puster, Richard L.: New Method for Deterining Heats of Cobustion of Gaseous Hydrocarbons. NASA TP-253, Singh, Jag J.; Chegini, Hoshang; and Mall, Gerald H.: A Siplified Method for Deterining Heat of Cobustion of Natural Gas. NASA TP-2682, Singh, Jag J.; Davis, Willia T.; and Puster, Richard L.: Proposed Fast-Response Oxygen Monitoring and Control Syste for the Langley 8-Foot High-Teperature Tunnel. NASA TP-228, McCracken, Dudley J.: Hydrocarbon Cobustion and Physical Properties. BRL Rep. No. 496, U.S. Ary, Sept (Available fro DTIC as AD ) 9. Miller, Richard D.; and Hertweck, Floyd R., Jr.: Analyses of Natural Gases, 982. Info. Circ. 8942, Bureau of Mines, U.S. Dep. of Interior. 0. Colean, Hugh W.; and Steele, W. Glenn, Jr.: Experientation and Uncertainty Analysis for Engineers. John Wiley & Sons, Inc.,

9 Burner Voluetric floweter PC Oxygen sensor Flow controller Flow controller Flow controller Controller Set point Pup Air Test gases Figure. Syste diagra. 7

10 Figure 2. Experiental setup. L

11 Σ f i H i, kcal/ol Methane ole fraction of hydrocarbon coponent Figure 3. Variation in calculated heat of cobustion Σf i H i with ethane ole fraction of hydrocarbon coponent for 265 natural gas well saples. 9

12 Σ f i H i /A Methane ole fraction of hydrocarbon coponent Figure 4. Variation of Σf i H i /A with ethane ole fraction of hydrocarbon coponent for 265 natural gas well saples. 0

13 Frequency Σ f i H i /A Figure 5. Distribution of Σf i H i /A for 265 natural gas well saples. Mean = 26.55; Standard deviation = 0.06.

14 Ethane data (5 pts.) Heat of cobustion, H, kcal/ol ±2 (SEE) H est = E Methane data (9 pts.) percent CH percent N 2 (9 pts.) Flow data, E, counts/3 in Figure 6. Suary of experiental data and regression results. 2

15 For Approved REPORT DOCUMENTATION PAGE OMB No Public reporting burden for this collection of inforation is estiated to average hour per response, including the tie for reviewing instructions, searching existing data sources, gathering and aintaining the data needed, and copleting and reviewing the collection of inforation. Send coents regarding this burden estiate or any other aspect of this collection of inforation, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Inforation Operations and Reports, 25 Jefferson Davis Highway, Suite 204, Arlington, VA , and to the Office of Manageent and Budget, Paperwork Reduction Project ( ), Washington, DC AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED March 996 Technical Paper 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS On-Line Measureent of Heat of Cobustion of Gaseous Hydrocarbon Fuel Mixtures WU AUTHOR(S) Danny R. Sprinkle, Sushil K. Chaturvedi, and Ali Kheireddine 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) NASA Langley Research Center Hapton, VA PERFORMING ORGANIZATION REPORT NUMBER L SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) National Aeronautics and Space Adinistration Washington, DC SPONSORING/MONITORING AGENCY REPORT NUMBER NASA TP SUPPLEMENTARY NOTES Sprinkle: Langley Research Center, Hapton, VA; Chaturvedi and Kheireddine: Old Doinion University, Norfolk, VA. 2a. DISTRIBUTION/AVAILABILITY STATEMENT 2b. DISTRIBUTION CODE Unclassified Unliited Subject Category 35 Availability: NASA CASI (30) ABSTRACT (Maxiu 200 words) A ethod for the on-line easureent of the heat of cobustion of gaseous hydrocarbon fuel ixtures has been developed and tested. The ethod involves cobustion of a test gas with a easured quantity of air to achieve a preset concentration of oxygen in the cobustion products. This ethod involves using a controller which aintains the fuel (gas) voluetric flow rate at a level consistent with the desired oxygen concentration in the cobustion products. The heat of cobustion is deterined fro a known correlation with the fuel flow rate. An on-line coputer accesses the fuel flow data and displays the heat of cobustion easureent at desired tie intervals. This technique appears to be especially applicable for easuring heats of cobustion of hydrocarbon ixtures of unknown coposition such as natural gas. 4. SUBJECT TERMS Natural gas; Natural gas energy content; Heat of cobustion; Hydrocarbon cobustion; Cobustion physics; Cobustion control; Enthalpy of cobustion; Theral energy easureent; Cobustion calorietry 7. SECURITY CLASSIFICATION OF REPORT 8. SECURITY CLASSIFICATION OF THIS PAGE 9. SECURITY CLASSIFICATION OF ABSTRACT Unclassified Unclassified Unclassified 5. NUMBER OF PAGES 6. PRICE CODE 20. LIMITATION OF ABSTRACT NSN Standard For 298 (Rev. 2-89) Prescribed by ANSI Std. Z A03