ISSN 0040-5795, Theoretical Foundations of Chemical Engineering, 2017, Vol. 51, No. 1, pp. 12–26. © Pleiades Publishing, Ltd., 2017. Original Russian Text © V.A. Kirillov, Yu.I. Amosov, A.B. Shigarov, N.A. Kuzin, V.V. Kireenkov, V.N. Parmon, Yu.V. Aristovich, M.A. Gritsay, A.A. Svetov, 2017, published in Teoreticheskie Osnovy Khimicheskoi Tekhnologii, 2017, Vol. 51, No. 1, pp. 15–30.

Experimental and Theoretical Study of Associated Petroleum Gas Processing into Normalized Gas by Soft Steam Reforming V. A. Kirillova, *, Yu. I. Amosova, A. B. Shigarova, N. A. Kuzinb, V. V. Kireenkova, V. N. Parmona, Yu. V. Aristovichc, M. A. Gritsayc, and A. A. Svetovc aBoreskov

Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia b Ltd UNICAT, Novosibirsk, 630090 Russia c BI Technology Ltd, St. Petersburg, 196084 Russia *e-mail: [email protected] Received April 26, 2016

Abstract⎯The article presents the results of experimental investigation and mathematical modeling of a new technology for converting associated petroleum gas to a normalized combustible gas that can be used in gas turbine and gas reciprocator power plants or, after removing part of the СО2, can be pipelined. The essence of the new technology is that the C2+ hydrocarbons contained in associated petroleum gas are converted by soft steam reforming into a gaseous fuel that consists mainly of methane and contains carbon dioxide and a small amount of hydrogen. This process increases the volume of the gas mixture and normalizes its heating value and Wobbe index to the standard characteristics of commercial natural gas (purified from СО2). The soft steam reforming technology has been tested on laboratory, pilot, and pre-commercial scales. A mathematical model has been developed for the process. A numerical analysis based on this model has demonstrated that, using this technology, it is possible to process associated petroleum gases varying widely in methane homologue concentrations in one tubular catalytic reactor. Keywords: associated petroleum gas, soft steam reforming, methane homologues, normalized gas, catalyst, process flowsheet, pilot unit, mathematical modeling, tubular reactor DOI: 10.1134/S0040579517010110

APG is 42–60 MJ/m3, so its application as a fuel would inevitably lead to engine overheating. In order to prevent this overheating of the APG-fired engine, the output of the power-generating unit is reduced by 25–30% relative to the nominal output corresponding to its type and size. Another problem in the use of APG in internal combustion engines is that APG contains macromolecular hydrocarbons that, depending on their concentration and gas pressure, condense below the dew point in the supply pipeline to form liquid drops at the inlet of the combustion chamber. As a consequence, the combustion of the fuel mixture takes place nonuniformly, is accompanied by tar formation in engine cylinders, increased coking, and detonation, and leads eventually to a premature engine failure [2]. Thus, for APG utilization as a fuel for local energy generation, it is necessary to develop a fairly simple and economically efficient technology for APG conversion to a normalized gas usable as a standard gaseous motor fuel in heat and electricity generation at petroleum extraction sites or, after СО2 removal, as a gas to be pipelined to the consumer. This technology, which is called soft steam reforming (SSR), was being developed for several years by the Boreskov Institute of

INTRODUCTION Associated petroleum gas (APG) is a by-product of oil recovery and contains a mixture of gaseous hydrocarbons from С1 to С8. APG is a valuable feedstock for obtaining various petrochemical products and for producing heat and electricity. At the same time, the problem of utilizing APG in Russia is a challenging one: according to the Ministry of Energy of the Russian Federation, 9.3 billion m3/yr of APG is flared in places of petroleum extraction [1]. This situation is due to the following circumstances. Firstly, much of the APG flared comes from small, low-pressure fields each producing less than 0.05 billion m3/yr of gas and the problem of utilizing APG from these fields is still unsolved. Secondly, direct use of APG as a motor fuel for local electricity production faces a number of serious difficulties. This is due to the fact that, depending on the oilfield, APG contains 30 to 80% methane and a considerable percentage of its heavy homologues. Gas-reciprocator-based power stations are equipped with internal combustion engines whose structural materials and heat transfer conditions are designed for operation on a fuel gas with a lower calorific value of at most 34–36 MJ/m3. The lower heating value of 12

EXPERIMENTAL AND THEORETICAL STUDY

Catalysis (Siberian Branch, Russian Academy of Science) and by BI Technology Ltd, an affilated company of the Institute [3, 4]. The essence of the technology is that the С2–С8 hydrocarbons contained in APG are catalytically converted under the action of steam to a gaseous motor fuel consisting mainly of methane. The methane present in APG is not involved in SSR, while its homologues turn into methane, hydrogen, and carbon dioxide, thus increasing the total volume of the gas mixture in proportion to the “fatty” hydrocarbon content of the initial APG. The normalized gas obtained after the removal of part of the СО2 has a methane number close to 100, a lower heating value of 31.8–32.5 MJ/m3 (20°C), and a higher Wobbe index of 42–47 MJ/m3 (20°C). These characteristics allow normalized gas, as distinct from APG, to be used without any limitations as a fuel for all types of energy generating units at petroleum extraction sites and to be pipelined. Experimental and theoretical studies of the simultaneous conversion of a С1–С4 hydrocarbon mixture under the action of air due to noncatalytic homogeneous reactions at pressures of 2.5–4.0 MPa and temperatures of 450–740°C [5] demonstrated a considerable decrease in the proportion of the С2+ fraction and an increase in the proportion of methane. Much earlier, in the 1970s–1980s, there were studies on natural gas purification from methane homologues (mainly from ethane) with steam at the Gas Institute of Ukraine [6–10]. Experiments were carried out on an ethane–natural gas model mixture over methanation catalysts containing about 50% nickel at 200– 400°C, a stoichiometric water excess coefficient of 2– 6, and a volumetric space velocity of 1400 h–1. Note that the present-day methanation catalysts of the NIAP series contain no more than 35% nickel and are equally active. This effect is attained by using an improved catalyst preparation procedure [11]. In more recent works [12–14], there were studies on low-temperature steam reforming of a 20–33% propane + methane mixture over nickel–chromium catalysts [12], on diesel fuel pre-reforming over the commercial nickel-containing catalysts NIAP-18 (Ni, 15 wt %; CaO, 8 wt %; Al2O3, 74.4 wt %), NIAP-12-05 (Ni, 48 wt %; Cr2O3, 27 wt %), NIAP-07-01 (NiO, 36 wt %), and NIAP-07-05 (NiO, 38 wt %; Cr2O3, 12 wt %) [13], and on the conversion of a 20% C2H6 + 10% C3H6 + methane mixture over Ni/Ce/Zr/La and Rh/Ce/Zr/La catalysts [14] into methane–hydrogen mixtures. Unfortunately, none of these works went beyond laboratory-scale tests. Thus, further theoretical and experimental studies are necessary to develop a technology for APG conversion to normalized gas. Here, we report the results of laboratory-scale theoretical and experimental studies carried out in 2009– 2015 at the Boreskov Institute of Catalysis on model mixtures in order to develop scientific foundations for the SSR technology, the results of pilot tests of the

13

SSR technology on real APG compositions in 2011 at the Nyagan’-Gazpererabotka Co. (Talinka, Khanty– Mansi Autonomous Okrug—Yugra) and in 2012 at the RN-Yuganskneftegaz Co. (Khanty–Mansi Autonomous Okrug—Yugra) in collaboration with the Zapsibtekhnologii Co. (Tyumen), and the results of tests on a pre-commercial scale in 2015 at the Krapivinskoe oilfield (Gazpromneft’-Vostok Co., Tomsk) in collaboration with BI Technology Ltd (St. Petersburg). THERMODYNAMICS OF SSR REACTIONS The target reactions of the SSR process are those of each component of the С2+ fraction:

C nH 2n+ 2 + n − 1 H 2O 2 (1) n 3 1 1 + − n CH 4 + CO 2, n = 2,3,…, 8. = 4 4 The selectivity of reaction (1) may be decreased by the formation of an additional amount of СО2 and Н2, so, in the general case, it is more convenient to represent the SSR reaction scheme as the following steps: steam reforming,

C nH 2n+ 2 + 2n H 2O → n CO 2 + (3n + 1)H 2, ; n = 2, 3,…, 8 . reverse water gas shift reaction, СО 2 + Н 2 ⇔ CO + H 2O, ; 0 Δ H 298 = + 41 kJ/mol

(2)

(3)

СО2 methanation,

CO 2 + 4H 2 ⇔ CH 4 + 2H 2O, Δ H 298 = – 1 65 kJ /mo l 0

;

(4)

coking,

CH 4 ⇔ 2H 2 + C,  Δ H 298 = + 7 5 kJ/ mol, CnHm ⇒ nC + m/2H2, 0

2CO ⇔ C + CO 2,

(5) (6)

(7) 0 Δ H 298 = − 17 2 kJ/ mol. A specific feature of the SSR reactions is that they combine reforming and methanation reactions as endothermic steps (2) and (3) and exothermic step (4) with coking reactions (5)–(7). As a result of the occurrence of the exothermic and endothermic reactions, the overall heat of the reaction and the composition of the reforming products depend on process conditions. This conclusion is clearly suggested by the calculated thermodynamic equilibrium data presented in Table 1. The calculation as carried out by the relaxation method; more specifically, the equations of the SSR reactor model (see below) were numerically solved for a sufficiently long reactor such that a thermodynamic

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

14

KIRILLOV et al.

Table 1. Temperature dependence of the equilibrium composition of the gas resulting from SSR followed by water condensation (residual water, 1%; nitrogen, 0.7−0.8%, to balance; no С2+) Т, °C 300 325 350 375 400 425 450 475 500

Сout, vol %

qreac,

Н2

СН4

СО2

СО

MJ/Nm3 (APG)

2.28 3.25 4.50 6.02 7.82 9.86 12.1 14.6 17.4

85.9 84.9 83.4 81.7 79.5 77.1 74.4 71.3 67.8

9.75 9.91 10.0 10.2 10.3 10.4 10.3 10.0 9.60

0.03 0.06 0.13 0.25 0.41 0.74 1.40 2.26 3.40

0.414 0.386 0.343 0.285 0.211 0.115 − 0.008 − 0.168 − 0.367

equilibrium is practically reached at its outlet at a preset temperature, a fixed pressure of 0.2 MPa, and an inlet H2O/С2+ molar ratio of 0.7 (which was calculated only for methane homologues) for the following APG composition (mol %): nitrogen, 1.44; carbon dioxide, 0.81; methane, 68.26; ethane, 11.13; propane, 11.85; butanes, 4.81; pentanes, 1.28; hexanes, 0.33; heptanes, 0.075; octanes, 0.015. The coking reactions (5)–(7) were left out of consideration. Figure 1 plots the hydrogen and methane concentrations (on dry gas basis) at thermodynamic equilibrium as a function of temperature and Н2О/С2+ molar ratio for SSR at 0.4 MPa involving typical “light” APG (mol %): methane, 71.64; ethane, 10.56; propane, 9.10; butanes, 4.51; pentanes, 1.35; hexanes, 0.73; heptanes, 0.16; octanes, 0.03; СО2, 0.61; nitrogen, 1.31. It follows from Fig. 1 that, by varying the temperature and Н2О/С2+ ratio, it is possible to efficiently control the hydrogen and methane concentrations in the APG reforming products. For maximizing the methane yield (and minimizing the hydrogen and carbon dioxide yields), SSR should be conducted at 250–350°C and an H2O/С2+ molar ratio of 0.50–0.70. A thermodynamic analysis of SSR in the selected range of process conditions, with reactions (5)–(7) taken into account, demonstrated the possibility of coke formation under these conditions. Experimental data on the reforming of natural gas [14, 15] and a propane–butane mixture [16] indirectly confirm that the tendency to coking strengthens with an increasing concentration of methane homologues and with a decreasing Н2О/С ratio. Note that the kinetic rate of coke formation on nickel catalysts was observed [16, 17] to decrease dramatically with a decreasing nickel particle size.

Qlow, MJ/m3 (without СО2) 32.5 32.2 31.9 31.5 31.0 30.5 29.9 29.2 28.4

EXPERIMENTAL SSR catalysts. The specific conditions under which SSR is conducted made it necessary to develop a catalyst that would be highly resistant to coking, and would be sufficiently active at 300–340°C. Developing a catalyst, we took into account that, in the future, it would be necessary to produce the catalyst in rather large amounts and the catalyst should be inexpensive. For this reason, we initially carried out a screening of commercial nickel and nickel–chromium catalysts of the NIAP series. Comparative tests were performed using a procedure reported in Ref. [12]. Before the comparative tests of catalyst samples, the samples were reduced with a 5 vol % hydrogen + nitrogen mixture [18]. The tests demonstrated that, with this activation method, H2, %

CH4, % 90

20 3 15

85

2 80

1 10

1

75 2

5 0 250

70

3 300

350 T, °C

60 450

400

Fig. 1. Equilibrium concentrations of hydrogen and methane resulting from SSR at a total pressure of 0.4 MPa as a function of temperature for inlet Н2О/С2+ molar ratios of (1) 2, (2) 1, and (3) 0.7.

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY

all of the catalyst samples show nearly equal SSR activities. As follows from the literature [16, 17], the steam reforming reaction is structure-sensitive; that is, it depends on the nickel particle size. The size of the nickel particles largely determines whether the reaction proceeds as steam reforming (2), methanation (4), or hydrocracking, the latter showing itself as coke formation. For example, at H2O/C2+ = 0.76 mol/mol and a nickel particle size of 20 nm, the coking rate in methane steam reforming is 0.001 g of carbon per gram of catalyst per hour [17], while at a particle size of 4–5 nm, the coking rate is approximately one order of magnitude lower. It was reported [16] that, as the nickel particle size is increased from 5.6 to 12.3 nm, the steam reforming of a methane and propane– butane fraction mixture switches to hydrocracking and this markedly enhances coke formation. Therefore, an optimal catalyst preparation technology should ensure that the nickel particle size is no larger than 6 nm. This is achieved by mixing oxygen-containing nickel, chromium, and aluminum compounds with an oxygen containing compound of magnesium and by adding graphite or a graphite-like carbon material to the catalyst stock to produce the necessary porous structure [19]. This carbon material is a three-dimensional carbon matrix with a pore volume of 0.2–1.7 cm3/g that is formed by ribbon-shaped carbon layers 10–1000 nm in thickness and has a true density of 1.80–2.10 g/cm3 and a porous structure with a pore size maximum in the 20–200 nm range or a bimodal porous structure with an extra pore size maximum in the 4–20 nm range. The oxygen-containing magnesium compound can be added while mixing the oxygen-containing compounds of nickel, chromium, and aluminum or after drying and/or calcination of the catalyst stock. The most optimal magnesium-to-aluminum ratio ensuring a nickel particle size no larger than 6 nm is Mg/Al = 0.25 [16]. Using a procedure suggested in Ref. [19], we prepared, for further studies, magnesium oxide–doped catalysts containing 50% nickel oxide (KAT-1) and 40% nickel oxide (KAT-2). Investigation of coke formation. In addition to performing a screening of catalysts according to their activity, it was essential to detertmine the coking resistance of the catalysts. The amount of carbon deposited on a catalyst during the SSR reaction was determined by burning the carbon in an oxygen-containing atmosphere and measuring the volume of the resulting carbon dioxide. Carbon brining was carried out on a TG 209 F1 Libra thermo-microbalance (NETZSCH). The СО2 concentration at the thermo-microbalance outlet was measured with a QMS-200 mass spectrometer. The experiments were performed both on a fresh catalyst and on a catalyst that had been on stream for 100 h. The fresh samples were found to contain 2.0 wt % carbon, which burned away in the temperature range

15

from 430 to 730°C. No additional carbon was detected on the catalyst that had been operated for 100 h. Coke formation on the doped catalyst KAT-1 was experimentally verified by examining the total carbon balance. For this purpose, we calculated the Н2/С molar ratio for the feed and for the reforming products obtained with the catalyst after 100-h-long testing. The (Н2/С)out/(Н2/С)in ratio was within the 0.95– 0.98 range. The close coincidence of the inlet and outlet ratios indicated that the carbon balance was nearly invariable, so there was no significant coking of the catalyst in our experiments. Pilot and pre-commercial SSR units. Experiments were initially carried out on model APG compositions using a pilot unit under benchmark conditions at the Boreskov Institute of Catalysis and then on real APG compositions at the Nyagan’-Gazpererabotka Co. (Talinka, Khanty–Mansi Autonomous Okrug— Yugra) in April 2011 and at the RN-Yuganskneftegaz Co. (Khanty–Mansi Autonomous Okrug—Yugra) in July 2012. Final tests were carried out on an SSR-300 pre-commercial unit (BI Technology Ltd, St. Petersburg) in December 2015 at the Krapivinskoe oilfield (Gazpromneft’-Vostok Co., Tomsk). The SSR pilot unit is schematized in Fig. 2. The unit, whose output capacity was up to 8 m3 of gas per hour, included the main elements modeling the operation of the pre-commercial unit to be designed: SSR reactor, burner assembly, steam generator–superheater, and condensing heat exchanger. The unit was portable, so it could be used to refine the operating regimes both on APG-imitating mixtures under benchmark conditions at the Boreskov Institute of Catalysis and on real APG under field conditions. Two SSR reactor variants, namely, an adiabatic one and a tubular one were fabricated. The adiabatic reactor had a diameter of 20 cm and a catalyst bed height of 80 cm, and the catalyst volume in it was 25 L. The tubular reactor consisted of seven tubes with an inner diameter of 20 mm, a wall thickness of 18 mm, and a bed height of 60 cm. A crushed catalyst with a particle size of about 2 mm was placed in the tubes. The total volume of the catalyst was 1 L. For compensating for the considerable heat loss and maintaining the catalyst temperature, hot flue gas with a controlled temperature, obtained by the two-stage catalytic combustion of natural gas in the burner assembly, was blown past the outer surface of the tubes. The burner assembly was fabricated according to the scheme presented in our earlier publications [20, 21]. The evaporating and steam superheating device was a tubular heat exchanger. Water to be evaporated was fed into the tubes, and the outer surface of the tubes, like that of the reactor, was heated with flowing flue gas. In the laboratory-scale experiments, the unit was additionally fitted with a SIEMENS stationary gas analyzer; in the field tests, it was equipped with a TEST-1 portable gas analyzer (BONER Co., Novosibirsk).

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

16

KIRILLOV et al.

e

a4

1

b1 b2 2

3 a3 d

c a2

4 a1

Fig. 2. Pilot SSR unit: (1) reactor, (2) burner assembly, (3) condensing heat exchanger, (4) evaporating heat exchanger, (a1) water for reaction, (a2) water after condensation, (a3) cold water into heat exchanger, (a4) hot water from heat exchanger, (b1) primary air + APG, (b2) secondary air, (с) APG, (d) SSR products, and (e) flue gas.

Benchmark experiments. These experiments were carried out in the tubular reactor using model gas mixtures imitating the APG composition. The conditions of some of the experiments and their results are specified in Table 2. The experiments were performed at Н2О/С2+ = 0.44–0.81 mol/mol, an inlet catalyst temperature of 319–344°C, and an outlet temperature of 287–310°C. Because it was impossible to activate the catalyst with a nitrogen–hydrogen mixture under the field conditions, we developed a catalyst activation

procedure involving the reaction mixture obtained by APG reforming with a considerable excess of water (Н2О/С2+ = 2–3 mol/mol) at increased temperatures of 400–420°C. Here, the outlet hydrogen concentration was up to 18–20%. Depending on the catalyst weight, activation was performed for several hours to tens of hours. This procedure was used before each new series of experiments. As was demonstrated by calculations, in dome experiments involving this activation procedure the hydrogen content of the reforming products for the KAT-1 catalyst was close to the equilibrium values listed in Table 1, while for KAT-2 the hydrogen content was well below the equilibrium values. These data indicated that the KAT-1 catalyst is more active because it contains more NiO than KAT-2 does. This is observed both for the methanation stage (4) and for the reforming stage (Table 2, experiments 3–6). On the whole, the experimental data presented in Table 2 confirm the results of thermodynamic analysis suggesting that, as the Н2О/С2+ ratio and temperature are decreased, the hydrogen concentration in the reforming products leaving the reactor decreases and the methane concentration increases. Pilot SSR tests under field conditions. Experimental testing of the SSR technology was carried out using the pilot unit schematized in Fig. 2, whose gas throughput capacity was 8 m3/h for real APG, at the Nyagan’-Gazpererabotka Co. (Talinka, Khanty– Mansi Autonomous Okrug—Yugra) in 2011 and at the RN-Yuganskneftegaz Co. (Khanty–Mansi Autonomous Okrug—Yugra) in 2012. In the experiments performed in Talinka, APG had the following composition (vol %): methane, 70.4; inert components (nitrogen and carbon dioxide), 7; “fatty” homologues, 22.6. The SSR product mixture had the following composition (vol %): methane, 86.2; hydrogen, 3.0; inert components, 13.2; “fatty” homologue, 0.57. The experimental studies were carried out using the KAT-1 catalyst and the adiabatic reactor. The pilot unit was integrated with an MTES-30 gas-fired reciprocatingengine unit (SINTUR-NT Co., Nizhny Tagil) with a maximum output power of 30 kW. The purpose of these experiments was to compare the performance of the gas-fired reciprocating-engine electricity generator operating on the initial APG to the performance of the same unit operating on normalized gas (APG reforming products). The results of these studies were reported in an earlier work [4]. It was found that use the fuel resulting from the SSR of APG raised the power output capacity of MTES-30 by 21% to bring it close to the nominal capacity of the energy generating unit, reduced the СО, СН, and О2 concentrations in the exhaust gas, enhanced the fuel combustion efficiency, and improved the dynamic characteristics of the engine. The experiments at the RN-Yuganskneftegaz facilities were performed to see whether the SSR technology can be used to convert APG to commercial-grade

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY

17

Table 2. Results of the benchmark experiments at the Boreskov Institute of Catalysis Т, °C

Gin Entry

gas, H2O, Н2О/С2+ inlet L/min g/h

1*)

4.00 200

0.51

2*)

5.25 100

0.50

3**)

10.0

290

0.81

4**)

10.0

210

0.59

5**)

13.3

210

0.44

6**)

13.3

220

0.46

344

outlet 310

Сout, vol % (on dry gas basis) CO

CO2

H2

C2

C3

C4

3.60

0.04

0.03

0.00

81.94

3.06

0.00

0.05

0.01

87.00

2.23

0.17

0.09

0.01

9.14 88.56

2.02

0.00

0.18

0.04

0.42 15.71 80.20

Calculation → 0.15 14.78 334

292

0.09

Calculation → 0.06 318

287

CH4

9.53

0.06

9.07

74.99 11.83

0.57

2.70

0.50

0.13

9.14

74.71 12.03

0.76

2.34

0.89

0.07

9.39

75.81 11.64

0.47

2.05

0.38

9.21 76.38 11.23

0.57

1.76

0.67

0.20

8.49 83.09

5.85

0.44

1.55

0.28

Calculation → 0.21

8.44 83.48

5.48

0.45

1.41

0.53

9.91

82.5

5.43

0.43

1.39

0.26

8.54 83.78

5.28

0.42

1.29

0.49

Calculation → 323

292

Calculation → 0.18 328 319

300 298

0.13

Calculation → 0.20

GHSV, h–1 (dry gas)

Х, %

629

99.6

826

98.6

600

75

600

81

800

84

800

86

* Experimental conditions for entries 1 and 2: catalyst KAT-1 (381 cm3); inlet gas composition in entry 1 (vol %): СО2 = 0.19, СН4 = 35.62, С2Н6 = 1.71, С3Н8 = 50.86, C4Н10 = 11.62; inlet gas composition for entry 2 (vol %): СО2 = 0.36, СН4 = 73.80, С2Н6 = 2.87, С3Н8 = 18.42, C4Н10 = 4.55. ** Experimental conditions for entries 3−6: catalyst KAT-2 (1000 cm3); inlet gas composition (vol %): СО2 = 0.43, СН4 = 75.25, С2Н6 = 4.62, С3Н8 = 14.3, C4Н10 = 5.43.

natural gas. These experiments were carried out on the same pilot unit (Fig. 2) but with the use of the tubular reactor. The SSR unit was additionally fitted with a liquid-phase alkali-based absorption block for СО2 removal from the product. Testing data (Table 3) indicate that the addition of the СО2 removal and water vapor condensation stage makes it possible to obtain normalized gas that practically meets the Gazprom standard 089-2010 and the USSR State Standard GOST 5542-87. It was also demonstrated that SSR a 37% relative increase in the volume of the combustible components of the mixture. Note that the composition of the reformed is almost the same as the gas composition corresponding to the thermodynamic equilibrium at 0.12 MPa, Тout = 305°C, and Н2О/С2+ = 0.8; in particular, the hydrogen concentration is 3.5% and is in agreement with the benchmark testing data obtained at the Boreskov Institute of Catalysis (Table 2). Thus, pilot-scale tests on real APG compositions demonstrated that the SSR technology provides means to convert methane homologues contained in APG of various compositions both to a gaseous motor fuel and to commercial-grade natural gas. This made it possible to proceed to pre-commercial testing of the technology. Pre-commercial testing of the SSR technology. The SSR technology was tested on an SSR-300 pre-commercial unit with an APG throughput capacity of up to

300 m3/h using the KAT-2 catalyst in December 2015 at the Krapivinskoe oilfield (Gazpromneft’-Vostok Co., Tomsk). The purpose of that work was to test the process flowsheet and the technical solutions used in the design of the SSR-300 pre-commercial unit. The flow diagram of the unit is presented in Fig. 3, and its photograph is shown in Fig. 4. The main blocks of the pre-commercial units are the following: APG pretreatment and supply block; heat exchanger HX-1000; aerial cooler HX-4000; separator V-5000; mixer MIX-1100; catalytic SSR reactor HX-2000; evaporator HX-3000/1; steam superheater HX-3000/2; deaerator V-6000; pumps P-3100, P3200, P-6100, and P-6200; chemical water treatment block CWT; tank T-7000. The pre-commercial unit was operated in the following way. The feedstock (APG) had the following composition (vol %): methane, 71.70; ethane, 7.81; propane, 7.46; isobutene, 1.43; n-butane, 2.50; isopentane, 0.63; n-pentane, 0.65; isohexane, 0.50; nhexane, 0.18; carbon dioxide, 4.63; nitrogen, 2.51. The APG, whose pressure was 0.3−0.5 MPa, was fed at a volumetric fl ow rate of 100−300 m3/h into the tube space of the heat exchanger HX-1000, where it was heated to the working temperature, 250−280°С. Next, in the mixer MIX-1100 it was mixed with superheated steam and was then fed into the catalytic SSR reactor HX-2000, where the steam–gas mixture was converted to normalized gas. After being successively

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

18

KIRILLOV et al.

Table 3. Results of the pilot tests at the RN-Yuganskneftegaz Co. Сout, vol % (on dry gas basis) Cin, vol % experiment

calculation experiment (after СО2 absorber)

СН4

78.75

86.37

С2Н6

3.89

0.01

0.039

0.01

С3Н8

6.62

0.03

0.066

0.03

С4Н10

4.92

0.06

0.049

0.06

С5Н12

1.99

0.06

0.02

0.06

С6Н14

0.61

0.03

0.006

0.03

С7Н16

0.16

0.01

0.002

0.01

С8Н18

0.027

0.02

0.0

0.02

СО2

0.88

8.40

8.71

0.95

СО

0.00

0.03

0.05

0.03

N2

2.15

1.45

1.45

1.57

H2

0.00

3.52

3.46

3.76

86.13

93.45

Qlow, MJ/m3

43.8

29.4

29.3

31.8

Wo, MJ/m3

54.1

41.3

41.0

47.4

−

98.1

98.3

−

X, % Increase in the dry volume of the mixture

1.00

1.48

1.48

1.37

Increase in the volume of combustible gases

1.00

1.375

1.37

1.375

cooled in the heat exchanger HX-1000 and aerial cooler HX-4000, the normalized gas was directed to the separator V-5000, where it was separated from excess condensed water, and then left the unit. Water from the separator V-5000 was recycled into the deaerator, where it was prepared for returning into the process. To prevent hydrate formation, isopropanol was sprayed into the fuel gas line. Industrial water supplied to the unit was subjected to chemical treatment in the CWT block. The chemically purified water from the receiver tank E-7000 was transported by the pumps P-6100 and P-6200 to the deaerator V-6000, where it was heated to 90−95°C with superheated steam to remove excess oxygen. The water pretreated in this way was delivered with the pumps P-3100 and P-3200 to the electric evaporator HX-3000/1 and then to the electric steam superheater HX-3000/2, where the steam was heated to the working temperature, 280−300°С. The hot steam was mixed with the gas in the mixer MIX-1100. An electric heater was used to carry out the starting heating of the reactor and to maintain the operating temperature in the intertubular space of the reformer. Electric heating was also used to generate and super-

heat steam. These technical solutions were used only in the pre-commercial SSR unit. In commercial units, a specialized heater operating on fuel gas is employed for the same purposes. Tests were performed in the operating regime corresponding to the optimum processing conditions. The reactor temperature was varied by controlling the heat carrier temperature by regulating the tubular electric heating element power. The following parameters were recorded in the steam reforming experiments: reactor temperature, reactor inlet and outlet pressures, volumetric APG flow rate at the inlet, and the composition of the reforming products at the outlet. The hydrogen and carbon dioxide concentrations in normalized gas at the separator outlet were continuously measured with a TEST-1 gas analyzer (BONER Co., Novosibirsk). No less frequently than once every 360 min, the initial APG and normalized gas were sampled to be analyzed for the concentration of each homological group of С2+ hydrocarbons on an FKhG-1M-2 portable chromatograph. Unfortunately, during the multiday experiments, the APG composition at the inlet of the unit was varying, even

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY

19

Normalized gas, 2–4 atmg НX-4000

APG, P = 3–5 atmg НX-1000

V-5000

MX-1100 V-6000

НX-3000/2

S

S S S

НX-2000

НX-3000/1

S

Water from well

P-3100, 3200

CWT E-7000 P-6100, 6200

Fig. 3. Flow diagram of the pre-commercial unit SSR-300.

if insignificantly, changing the reactant molar ratio and temperature at the reactor inlet. After the unit was brought to the operating conditions, the higher hydrocarbons-to-APG conversion at a process temperature of 310−315°C and an APG fl ow rate of 155−250 m3/h was 80−94%. The methane concentration in the commercialgrade gas was 79−82%, the hydrogen concentration in the conversion products was 6−8%, and the carbon dioxide concentration in the products was 7−8%. In case carbon dioxide was removed, the methane concentration could be increased to 90%. The unit was operated in the continuous automated mode for 6 days. The most representative data of the tests are listed in Table 4. Thus, the field experiments on APG gave anticipated results concerning the use of the SSR technology in APG conversion into a gaseous motor fuel and normalized gas.

MATHEMATICAL MODELING Mathematical model of the catalytic SSR reactor. Model equations were set up under the following assumptions: (1) We considered a unidimensional two-phase mathematical model of a tubular reactor with a granular catalyst bed inside the tubes and a liquid or boiling coolant flowing outside. Thermal calculations were performed using a three-temperature model and effective radial heat transfer coefficients. The independent variables of the model were the coolant temperature, the reacting gas temperature, and the mean catalyst temperature along the tube radius. Based on these quantities, we calculated the tube wall temperature, the catalyst temperature along the tube axis and near the wall, taking into account the additional heat transfer resistance in the near-wall zone. (2) A unidimensional model of single-phase (liquid) coolant flow in the intertubular space was considered.

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

20

KIRILLOV et al.

Fig. 4. Pre-commercial unit SSR-300.

(3) The axial dispersion of heat and matter was neglected; that is, a plug-flow reactor was considered. (4) The external diffusion limitations on the rate of СnНm steam reforming in the catalyst grain were taken into account in terms of analytical formulas for iso-

thermal grain efficiency in a first-order reaction with respect to the reactant. (5) Equations for the longitudinal transfer of the mixture components were solved in mass fractions to rigorously take into account the change in the volume

Table 4. Results of the experiments carried out on the pre-commercial unit SSR-300 (Тin = 310°C, Рin = 0.34 MPa) Experiment 1* Component

Сin, vol %

Сout, vol % (dry mixture) experiment

Methane Ethane Propane Isobutane Butane Isopentane Pentane Isohexane Hexane Hydrogen Nitrogen Carbon dioxide

71.70 7.81 7.46 1.43 2.50 0.63 0.65 0.50 0.18 0.00 2.51 4.63

79.52 1.26 0.95 0.17 0.25 0.05 0.05 0.02 0.01 8.00 1.94 7.60

Experiment 2** Хn, %

Сin, vol %

calculation 77.38 1.04 0.99 0.19 0.33 0.08 0.09 0.07 0.02 8.04 1.73 10.61

– 79.38 83.72 84.81 87.22 89.86 90.17 94.89 92.90 – – –

71.70 7.81 7.46 1.43 2.50 0.63 0.65 0.50 0.18 0.00 2.51 4.63

Сout, vol % (dry mixture)

Хn, %

experiment

calculation

81.77 0.94 0.87 0.15 0.24 0.05 0.05 0.02 0.01 6.70 1.93 7.10

78.53 0.88 0.84 0.16 0.28 0.07 0.07 0.03 0.02 6.49 1.73 10.64

– 84.25 84.73 86.27 87.43 89.61 89.93 94.76 92.73 – – –

* December 8, 2015. Conditions: APG flow rate, 169 m3/h. ** December 9, 2015. Conditions: APG flow rate, 178 m3/h. THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY

of the mixture as a result of reforming. To determine the reaction rate from the partial pressures of mixture components, the mass fractions in each step along the reactor were converted to mole fractions. (6) Kinetic relationships and their parameters for the steam reforming of all components of the С2+ fraction (ethane, propane, butane, etc.) were set to be the same. (7) A group of irreversible steam reforming reactions of components of the С2+ fraction (2), reversible shift reaction (3), and reversible methanation reaction (4) were considered. The model equations appear as follows (0 ≤ z ≤ L):

Gg

where 0 Eref = 112 kJ/mol, log10(k ref ) = 10.8 s–1.

(15)

2 ⎡ ⎤ P P W met = k met PH 2 ⎢1 − metCH 4 H2O 4 ⎥ , ⎣ K eq PCO2 (PH2 ) ⎦ −E 0 k met = k met exp ⎛⎜ met ⎞⎟ , ⎝ R gTc ⎠

(16)

where 0 Emet = 50 kJ/mol, log10(k met ) = 5.8 s–1.

∑

−1 −1 ηcat = 3 (tanh (ϕ) − ϕ ), ϕ exp(ϕ) − exp(−ϕ) tanh(ϕ) = , exp(ϕ) + exp(−ϕ)

(8)

d ⎛ k* ⎞ ϕ = p ⎜⎜ refref ⎟⎟ 2 ⎝ χDg ⎠

where

c gG g

dT g = FcatS b hg (Tc − T g ), dz

(9)

c fG f

dT f = At h f (TW − T f ). dz

(10)

⎛ 8 ⎞ (n) (n) Fcat ⎜ Δ H ref W ref + Δ H metW met + Δ H shW sh ⎟ ⎜ ⎟ (11) ⎝ n=2 ⎠ ef = FcatS b hg Tc − T g + At hr (Tc − TW ) ,

∑(

)

(

)

h f (TW − T f ) − hr (Tc − TW ) = 0, ef

−1

−1 −1

where hr = (hr + hW ) . ef

(12)

Inlet conditions (z = 0):

T g = T g 0, T f = T f 0, yi = yi 0.

(13)

Kinetic parameters for the steam reforming of each СnН2n + 2 component (n = 2, 3, …, 8) and for СО2 methanation were set according to the data obtained by mathematical processing of experimental kinetic data obtained for a propane–methane mixture with nickel–chromium catalysts in a laboratory-scale flow reactor [12]:

0 n = 2, 3, … 8, k ref = k ref

0.5

.

P P ⎛ ⎞ W sh = k sh PCO2 ⎜1 − shCO H 2O ⎟ , ⎝ K eq PCO2 PH 2 ⎠ (19) −E sh ⎞ 0 ⎛ k sh = k sh exp ⎜ ⎟. ⎝ R gTc ⎠ The temperature dependences of the equilibrium met sh constants K eq and K eq were set in the form of empirical correlation available from the literature [18]. The heat transfer coefficient hf for a liquid coolant flowing across the tube bundle was also set using empirical correlations [23]. The coefficient of heat transfer between the reacting gas and the granular catalyst bed, hg, was set using commonly accepted relationships for a granular bed under forced gas convection conditions [24]. The effective radial heat transfer coefficient for a tube with a granular catalyst was determined as follows [22]:

8λ r (20) . D The heat transfer coefficient in the wall zone of the tube was determined as follows [23]: hr =

hW =

(n)

(14)

(18)

The rate of the reverse shift reaction was set in a formal way so that the rate constant is sufficiently large for this reaction to proceed near its thermodynamic equilibrium:

W ref = ηcat k ref PC nH 2n + 2 , −E exp ⎛⎜ ref ⎞⎟ , ⎝ R gTc ⎠

(17)

The efficiency of the spherical catalyst grain, η cat , for the first-order reaction was determined using the following analytical formulas [22]:

dyi dz

⎛ ref 8 ⎞ (n) = Fcat ⎜ ν i,n W ref + ν imetW met + ν ishW sh ⎟ mi , ⎜ ⎟ ⎝ ⎠ n=2

21

NuW =

NuW λ g , d eq

(21)

0.8 0.33 0.09Re g Pr g .

For the radial thermal conductivity of the granular bed, we used the following empirical correlations,

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

22

KIRILLOV et al.

TC, °C 400

(а) T

380 360 340 320 300 280

A

260 240 0

0.5

1.0

2.0 z, m

(b)

XC2+, % 100 80

1.5

T

60 40 20

0

A

0.5

1.0

1.5

2.0 z, m

Fig. 5. Calculated profiles of the (a) catalyst temperature and (b) С2+ conversion in the tubular (T) and adiabatic (A) reactors at a fixed inlet temperature of 340°C for the SSR of “fatty” APG.

which took into account the convective component and heat radiation [23]:

λ r = λ g (1.5 + 0.1Re Pr) + λ rad , 1 − ε cat λ ef + ε catλ rad , rad = 1 + 1 λ cat λ rad λ rad = ε radσ radTc3d cat. ef

(22)

In the numerical calculation of model equations, we used a home-made program code implemented in the programming language C to realize a finite-difference method with a variable step size along the tube length for solving the transfer equations and for carrying out a umber of iterative procedures necessary for solving nonlinear model equations. Results of reconstruction of experimental data obtained on the pre-commercial unit. A comparison

between the data calculated using the mathematical model and the experimental data obtained on the precommercial unit is presented in Table 4. In the calculation, we assumed that the tube walls are ideally maintained at 310°C by the liquid coolant boiling in the intertubular space and took into account the radial and axial temperature gradients in the tube with granular catalyst bed, allowing for the near-wall thermal resistance inside the tube. The calculated and experimental concentrations of hydrogen, carbon dioxide, and С2+ methane homologues at the reactor outlet are in fairly good agreement, but note that the outlet gas composition differs significantly from the gas composition at thermodynamic equilibrium. For this reason, for attaining a better fit between the calculated and experimental data, we had to decrease the rate constants of С2+ reforming reactions (2) and particularly that of methanation (4). As was mentioned in the discussion of experimental data, this may be due to the difference between the activities of the KAT-1 and KAT-2 catalysts. Summarizing the results of the experiments and comparing these results with the data calculated using the model, we arrived at the conclusion that the mathematical model suggested here provides a quite adequate description for the performance of the SSR reactor under different conditions. Results of mathematical modeling of SSR. An important feature of the SSR technology is substantiation of choosing an adiabatic or tubular reactor for conducting the process. Consider the SSR of “fatty” APG (the APG composition and reforming conditions are presented below). In case the limiting temperature at the reactor outlet is 340°C (the necessity of this constraint for ensuring the minimum possible Qlow follows from thermodynamic estimates presented in Table 1), the inlet temperature for the tubular reactor with heat removed by the coolant taken into account should be set at a rather high level of 320°C. For the adiabatic reactor, because of the exothermicity of the overall SSR process, the inlet temperature should be decreased to the lowest possible level for the catalyst— 240°C. Under these conditions, the reactant conversion in the tubular reactor is 90% already at a bed height of 0.6 m and the conversion in the adiabatic reactor is only 60% (Fig. 5). Thus, the calculations demonstrate that, at the same inlet temperature, the tubular reactor is considerably superior to the adiabatic reactor in terms of output capacity in the SSR of APG. The next example of calculations presents an analysis of the possibility of carrying out the SSR of APGs with very different С2+ fraction contents in one tubular reactor. Suppose “light” APG has the following composition (vol %): methane, 71.60; ethane, 10.56; propane, 9.10; butanes, 4.51; pentanes, 1.35; hexanes, 0.73; heptanes, 0.16; octanes, 0.03; СО2, 0.65; nitrogen,

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY

23

Table 5. Calculated parameters of normalized gas resulting from the SSR of “light” and “fatty” APG in the tubular reactor and the removal of water and partial removal of carbon dioxide APG “Light” “Fatty”

Тout, °C

Н2

317 340

3.12 3.95

Сout, vol %

Vout,

СН4

СО

СО2

N2

Nm3/h

Qref, kW

MJ/m3

Wo, MJ/m3

94.87 94.65

0.08 0.10

1.03 1.00

0.91 0.30

3000 3000

209 428

32.0 32.1

47.8 48.1

1.31. Inlet flow rates: APG, 2089 m3/h; water, 921kg/h; coolant, 8617 kg/h. “Fatty” APG (vol %): methane, 22.61; ethane, 9.60; propane, 32.21; butanes, 23.50; pentanes, 6.81; hexanes, 3.05; heptanes, 1.05; octanes, 0.15; СО2, 0.25; nitrogen—0.77. Inlet flow rates: APG, 1173 m3/h; water, 1791 kg/h; coolant, 17712 kg/h. SSR reactor parameters: pressure, 0.4 MPa (absolute); inner diameter of the tubes with a granular catalyst, 80 mm; grain size, 5 mm; transverse tube spacing in the tube bundle, 119 mm (for a triangular tube arrangement); height of the tube bundle, 1.5 m; number of tubes in the bundle, 217; baffles ensuring a crossflow of the liquid coolant (high-temperature organic heat transfer medium) over the tubes are placed at 0.1-m intervals along the height of the tube bundle; cocurrent flows of the reaction mixture and coolant. The results of SSR calculations for these parameters are presented in Table 5 and Fig. 6. Figure 6a illustrates the typical thermal regime of the tubular SSR reactor. Because of the exothermicity of the process, which is due to the methanation reaction, the maximum catalyst temperature is observed at the tube axis. As the reforming process and heat transfer to the coolant come to completion, the catalyst temperature decreases. The largest radial temperature drop is observed within the granular bed (D/dp = 16), but the contribution from the near-wall thermal resistance is significant as well. Note that the total amount of heat released in the SSR of the “fatty” gas is more than 2 times larger than the amount of heat released in the SSR of the “light” gas (Table 5). The proportional increase of the coolant flow rate counterbalances this increase in exothermicity, and, as a result, the outlet temperatures for the “light” and “fatty” gases are relatively close. The С2+ conversion profiles for the “light” and “fatty” APGs are fairly similar (Fig. 6b). Thus, the above calculations demonstrate that by tuning the inlet conditions for both the “light” and “fatty” APG compositions, it is possible to obtain, in the same tubular reactor, practically equal yields of normalized gas with similar compositions and properties (Table 5). With an adiabatic reactor, this universality in terms of APG composition variation would obviously be impossible. Note that, in Table 5, the gas composition after SSR is close to the composition corresponding to the

Qlow,

thermodynamic equilibrium; that is, it is in agreement with the equilibrium calculations (Table 1, Fig. 1), considering the difference between the APG pressures and compositions. The calculated composition of normalized gas in Table 5 is close to the composition obtained in the pilot tests at the RN-Yuganskneftegaz facilities (Table 3). At the same time, there is a noticeable deviation from the equilibrium gas composition for the KAT-2 catalyst, which was employed in the TC, °C

(а)

400

1

380 360 340 2

320 300

3

280

4

260 240 0

0.25

0.50

0.75

1.00

1.25

(b)

XC2+ , %

1.50 z, m

100 80

H L

60 40 20

0

0.25

0.50

0.75

1.00

1.25

1.50 z, m

Fig. 6. (a) Calculated temperature profiles in the tubular reactor for the SSR of “light” APG: (1) catalyst at the tube axis, (2) catalyst near the tube wall, (3) tube wall, and (4) coolant. (b) Comparison of the C2+ conversions in the SSR of “light” (L) and “fatty” (H) APGs.

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

24

KIRILLOV et al.

benchmark tests at the Boreskov Institute of Catalysis (entries 3–6 in Table 2) and in the pre-commercial tests at the Krapivinskoe oilfield (Table 4). For example, the experimentally observed outlet hydrogen fraction, 6–8 vol %, is several times larger than the calculated equilibrium value. As was mentioned in the Experimental section, a possible cause of this discrepancy is the insufficient activity of KAT-2, as distinct from KAT-1, for which the composition of the reforming products is typically close to the equilibrium composition. CONCLUSIONS The benchmark test, pilot-scale and pre-commercial experimental studies and the results of mathematical modeling clearly demonstrate that the SSR technology is promising for APG processing into normalized gas that is similar in characteristics to commercial natural gas. This normalized gas can be used as a gaseous motor fuel at the APG extraction site or pipelined (after partial СО2 removal). Depending on the С2+ content of APG, the process results in a considerable increase in the volume of the combustible mixture. An important advantage of the SSR technology is that it provides means to convert APG with low and high С2+ contents into normalized gas in one tubular reactor by regulating the flow rates of APG, steam, and coolant. A specific feature of the SSR process is that it consists of a single stage and does not recycle the reacting gas, since the steam reforming and methanation reactions are combined and carried out over the same catalyst. These factors make up a significant competitive advantage of the SSR technology over the other APG processing technologies. We believe that, at present, there is no other commercially reasonable technical solution intended for small and low-pressure oilfields and aimed at obtaining a normalized gas fuel for local autonomous energy production. It is important that the gas resulting from the SSR of APG in the cooled tubular reactor followed by partial СО2 removal and drying meets the USSR State Standard GOST 5542-87 for town gases and Gazprom standard 089-2010 for gas pipelining in terms of the minimum lower heating value (31.8 MJ/m3) and higher Wobbe index range (41.2–54.5 MJ/m3). With the adiabatic reactor that is used in competing technologies [25], the same results it are difficult to achieve even for “light” APG because of the necessity of lowering the inlet temperature and, hence, an unreasonable increase in the amount of catalyst and are practically unattainable for “fatty” APG. ACKNOWLEDGMENTS The authors are grateful to V.A. Sobyanin, P.V. Snytnikov, V.D. Belyaev, and D.I. Potemkin (Boreskov Institute of Catalysis) for fruitful discussions of the work; A.V. Sazonov, R.R. Vergasov,

L.V. Saprykin, L.V. Saprykina, and D.A. Kruglikov (BI Technology) for providing pre-commercial testing data; L.V. Kozodoev, Director of Zapsibtekhnologii (Tyumen), untimely deceased, for his direct participation in organizing and carrying out the pilot tests at Yugra oilfields (Talinka, Nefteyugansk). NOTATION At Сin Сout cg, cf D dp

total perimeter of the cross section of all tubes, m composition of the initial model mixture or APG, vol % composition of the SSR products (on dry gas basis), vol % specific heat capacity of the reacting gas and coolant, J/(kg K) inner tube diameter, m catalyst grain diameter, m

Dg

effective diffusion coefficient of СnH2n + 2 in

frad

dimensionless correction factor accounting for heat transfer through reradiation from grains in the bed

Fcat Gg

total cross-sectional area of all tubes, m2 reaction mixture flow rate, kg/s

Gf

liquid coolant flow rate, kg/s

GHSV

gas hourly space velocity, h–1 coefficient of heat transfer between the reacting gas and the catalyst, W/(m2 K) coefficient of heat transfer between the coolant and the tube wall, W/(m2 K) near-wall heat transfer coefficient on the granular bed side, W/(m2 K) radial heat transfer coefficient in the tube with a catalyst, W/(m2 K) heat of steam reforming reactions of СnH2n + 2 (n = 2, 3, …, 8) yielding СО2 and Н2, J/mol

ef

the reaction mixture, m2/s Eref, Emet, Esh activation energies for reaction steps, J/mol

hg hf hW hr (n) Δ H ref

ΔHmet

heat of СО2 methanation, J/mol (СО2)

ΔHsh

heat of the reverse shift reaction, J/mol (СО2)

kref

rate constant of the steam reforming of СnH2n + 2, mol/(m3 s atm)

kmet, ksh

rate constants of the methanation and shift reactions, mol (СО2)/(m3 s atm)

met

K eq

sh K eq

thermodynamic equilibrium constant for the methanation reaction, atm2 dimensionless thermodynamic equilibrium constant for the reverse shift reaction

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

EXPERIMENTAL AND THEORETICAL STUDY L mi

length of the tubular reformer, m molar mass of a component of the reacting gas, kg/mol amount of heat released on passing from the initial state to the thermodynamic equilibrium, MJ/m3 (APG)

qreac

Qlow Qref Pi Rg Sb

СО2 methanation;

n

partial pressure of a reacting gas component, atm gas constant, J/(mol K)

r ref

component number in the С2+ fraction (n = 2, 3, …, 8): 2—ethane, 3—propane, and so on radial steam reforming of СnН2n + 2

sh

reverse shift reaction

Тg

reacting gas temperature, K

Тf

liquid coolant temperature, K

Vout

yield of the product mixture, m3/h

Wo

higher Wobbe index at 20 °C, MJ/m3 rate of the steam reforming of СnH2n + 2,

Wmet, Wsh

mol/(m3 (bed) s) rate of the methanation and reverse shift reactions, mol (СО2)/(m3 (bed) s)

X Xn

conversion of the С2+ fraction, % CnH2n + 2 conversion, n = 2, 3, …, 8, %

yi

dimensionless mass fraction of a component

z

ηcat

coordinate along the reformer length, m dimensionless void fraction of the granular bed dimensionless degree of blackness of the catalyst grain surface dimensionless catalyst grain efficiency

λ cat

thermal conductivity of grains, W/(m K)

λ cat

thermal conductivity of the granular bed due to radiation, W/(m K) radial thermal conductivity of the catalyst bed, W/(m K) dimensionless stoichiometric coefficients

ε cat ε rad

λr ν

σ rad χ

thermal radiation constant, W/(m2 K4) dimensionless permeability of the catalyst grain

SUBSCRIPTS AND SUPERSCRIPTS in out c, cat

CH 4 ,H 2O, CO,H 2,CO 2 , N 2 } met

ТW

(n) W ref

reacting gas coolant reacting gas components, i = {C nH 2n+ 2 (n = 2, ...8),

lower heating value at 20 °C, MJ/m3 total amount of heat rekeased in SSR, kW

specific surface area of the catalyst bed, m–1 catalyst temperature averaged over the tube radius, K tube wall temperature, K

Tc

g f i

25

reformer inlet reformer outlet catalyst

REFERENCES 1. http://minenergo.gov.ru/node/1156. 2. Rybakov, B.A., Burov, V.D., Rybakov, D.B., and Trushin, K.S., Specific features of the combustion of associated petroleum gas in gas-turbines, Turbiny Dizeli, 2008, no. 3, p. 2. 3. Snytnikov, P.V., Kirillov, V.A., Sobyanin, V.A., Belyaev, V.D., Kuzin, N.A., Kireenkov, V.V., Amosov, Yu.I., Polyanskaya, T.V., Popova, M.M., and Potemkin, D.I., RF Patent 2442819, 2012. 4. Kozodoev, L.V., Kuzin, N.A., Amosov, Yu.I., Kirillov, V.A., Sobyanin, V.A., and Kireenkov, V.V., New technology for processing associated petroleum gas in places of its recovery, Prom–st. Ekol. Severa, 2011, no. 11, p. 40. 5. Arutyunov, V.S., Rudakov, V.M., Savchenko, V.I., and Sheverdenkin, E.V., Relative conversion of lower alkanes in their simultaneous partial gas-phase oxidation, Theor. Found. Chem. Eng., 2005, vol. 39, no. 5, p. 487. 6. Meshchenko, N.T. and Veselov, V.V., Catalysts for removing methane homologues from natural gas, Khim. Tekhnol., 1972, no. 2, p. 47. 7. Meshchenko, N.T., Veselov, V.V., Shub, F.S., and Temkin, M.I., Kinetics of low-temperature ethane steam reforming over a nickel–chromium catalyst, Kinet. Katal., 1977, vol. 18, no. 4, p. 962. 8. Meshchenko, N.T. and Veselov, V.V., Kinetics of lowtemperature hydrocarbon reforming in a flow reactor, Khim. Tekhnol., 1977, no. 5, p. 41. 9. Reikhert, A.L., Sennikova, A.I., Krasnopevtsev, Yu.F., Tsvetkov, V.F., Meshchenko, N.T., and Nazarov, R.K., Catalytic stabilization of associated petroleum gas to be subjected to high-temperature steam oxygen reforming, in Sbornik trudov Instituta Gaza AN USSR (Collection of Papers of the Gas Institute of the USSR Academy of Sciences), Kiev: Naukova Dumka, 1979, p. 45. 10. Kervalishvili, Z.Ya., Mekhradze, D.V., Sagareishvili, Ts.Sh., Gulua, D.P., Akinin, V.A., Veselov, V.V., and Meshchenko, N.T., Testing of a new process for steam purification of associated petroleum gas from methane homologues on a pilot plant, in Sbornik trudov Instituta Gaza AN USSR (Collection of Papers of the Gas Institute of the USSR Academy of Sciences), Kiev: Naukova Dumka, 1979, p. 52.

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Vol. 51

No. 1

2017

26

KIRILLOV et al.

11. Efremov, V.N., Golosman, E.Z., Kashinskaya, A.V., and Tesakova, G.M., Methanation of carbon oxides over new, highly active, industrial catalysts of the NKM-7 (NIAP-07-07) series, Khim. Prom–st., 2012, no. 11, p. 5. 12. Zyryanova, M.M., Snytnikov, P.V., Shigarov, A.B., Belyaev, V.D., Kirillov, V.A., and Sobyanin, V.A., Low temperature catalytic steam reforming of propane– methane mixture into methane-rich gas: Experiment and macrokinetic modeling, Fuel, 2014, vol. 135, p. 76. 13. Kirillov, V.A., Shigarov, A.B., Amosov, Yu.I., Belyaev, V.D., and Urusov, A.R., Diesel fuel pre-reforming into methane–hydrogen mixtures, Theor. Found. Chem. Eng., 2015, vol. 49, no. 1, p. 30. 14. Angeli, S.D., Pilitsis, F.G., and Lemonidou, A.A., Methane steam reforming at low temperature: Effect of light alkanes presence on coke formation, Catal. Today, 2015, vol. 242, p. 119. 15. Sperle, T., De Chen, Lodeng, R., and Holmen, A., Pre-reforming of natural gas on Ni catalyst: Criteria for carbon free operation, Appl. Catal., A, 2005, vol. 282, p. 195. 16. Tan, M., Wang, X., Shang, X., Zou, X., Lu, X., and Ding, W., Template-free synthesis of mesoporous γalumina supported Ni–Mg oxides and their catalytic properties for pre-reforming liquefied petroleum gas, J. Catal., 2014, vol. 314, p. 117. 17. Christensen, K.O., Chen, D., Lodeng, R., and Holmen, A., Effect of support and Ni crystal size on carbon

18. 19. 20.

21.

22. 23. 24.

25.

formation and sintering during steam methane reforming, Appl. Catal., A, 2006, vol. 314, p. 9. Spravochnik azotchika (Nitrogen Engineer’s Handbook), Mel’nikov, E.Yu., Ed., Moscow: Khimiya, 1986. Snytnikov, P.V., Kirillov, V.A., Amosov, Yu.I., and Sobyanin, V.A., RF Patent 2568810, 2015. Kirillov, V.A., Kireenkov, V.V., Kuzin, N.A., Samoilov, A.V., and Shigarov, A.B., Catalytic external combustion engine, Theor. Found. Chem. Eng., 2015, vol. 49, no. 4, p. 375. Kirillov, V.A., Shigarov, A.B., Samoilov, A.V., Kuzin, N.A., Kireenkov, V.V., and Ivanov, D.A., Development of a catalytic heating system for external combustion engines, Theor. Found. Chem. Eng., 2016, vol. 50, no. 1, p. 1. Lapidus, N. and Amundson, N.R., Chemical Reactor Theory: A Review, New Jersey: Prentice-Hall, 1977. Spravochnik po teploobmennikam (Handbook on Heat Exchangers), Moscow: Energoatomizdat, 1987, vol. 1. Shigarov, A.B. and Kirillov, V.A., Modeling of membrane reactor for steam methane reforming: From granular to structured catalysts, Theor. Found. Chem. Eng., 2012, vol. 46, no. 2, p. 97. Kaila, R. and Jansson, P., Converting low quality gas into a valuable power source, Wärtsilä Tech. J., 2013, vol. 1, p. 61.

THEORETICAL FOUNDATIONS OF CHEMICAL ENGINEERING

Translated by D. Zvukov

Vol. 51

No. 1

2017