METHOD FOR ELECTROCHEMICAL HYDROGEN SEPARATION FROM NATURAL-GAS PIPELINES

20230111285 · 2023-04-13

    Inventors

    Cpc classification

    International classification

    Abstract

    The present invention comprises a one-stage membrane process for electrochemical separation of hydrogen from natural gas streams in a pipeline (1) having a positive pressure in the range from 50 mbar to 100 bar, having the following process steps: (i) a gas substream (2) is drawn off from the natural gas stream in a pipeline (1) without any change in the gas composition, where the mass flow rate of the gas substream is adjusted depending on the hydrogen content in the natural gas stream (1) such that a depletion level of 0.65 to 0.975 is established in the case of a hydrogen concentration of <10% by volume and a depletion level of 0.55 to 0.925 in the case of a hydrogen concentration of >10% by weight, where the depletion level is defined as the quotient of the desired molar H2 product stream (6) and the molar H2 reactant flow rate in the gas substream at the inlet of the membrane unit (2), (ii) this gas substream (2) is compressed (3) upstream of a membrane unit (5), (iii) this gas substream is heated to 100 to 250° C. either upstream of the membrane unit or in the membrane unit, and this gas substream is supplied with water (4) upstream of the membrane unit and/or on the permeate side of the membrane unit (4a), such that the water loading is between 0.005 and 0.2 mol of water/mol of natural gas, (iv) this gas substream is sent to an electrochemical membrane unit in which hydrogen is separated off as permeate (6a) at a temperature of 100 to 250° C., (v) the retentate (8) from the membrane unit is recycled into the natural gas stream, sent to a chemical utilization and/or used as fuel.

    The present invention further comprises a method of ascertaining the optimized gas substream which is drawn off from a pipeline that conducts natural gas and hydrogen in order to separate hydrogen from this gas substream in an electrochemical membrane unit.

    Claims

    1.-12. (canceled)

    13. A one-stage membrane process for electrochemical separation of hydrogen from natural gas streams in a membrane unit, having the following process steps: (i) drawing a gas substream off from a natural gas stream in a pipeline having a positive pressure in the range from 50 mbar to 100 bar without any change in the gas composition, where the mass flow rate of the gas substream is adjusted depending on the hydrogen content in the natural gas stream such that a depletion level of 0.65 to 0.975 is established in the case of a hydrogen concentration of <10% by volume and a depletion level of 0.55 to 0.925 in the case of a hydrogen concentration of >10% by weight, where the depletion level is defined as the quotient of the desired molar H2 product stream and the molar H2 reactant flow rate in the gas substream at the inlet of the membrane unit, (ii) compressing the gas substream upstream of the membrane unit, (iii) heating the gas substream to 100 to 250° C. either upstream of the membrane unit or in the membrane unit, and this gas substream is supplied with water, with supply of water to the gas substream either upstream of the membrane unit and/or to the permeate side of the membrane unit (4a), such that the water loading is between 0.005 and 0.2 mol of water/mol of natural gas, (iv) supplying the gas substream to the electrochemical membrane unit and hydrogen is separated off as permeate at a temperature of 100 to 250° C., (v) recycling the retentate from the membrane unit into the natural gas stream in the pipeline, sent to a chemical utilization and/or used as fuel.

    14. The process according to claim 13, wherein the pipeline has a positive pressure of 100 mbar to 10 bar.

    15. The process according to claim 13, wherein the pipeline has a positive pressure of 100 mbar to 2 bar.

    16. The process according to claim 13, wherein the hydrogen content of the natural gas streams in the pipeline is 0.1% to 20% by volume.

    17. The process according to claim 13, wherein the hydrogen content of the natural gas streams in the pipeline is 0.5% to 5% by volume.

    18. The process according to claim 13, wherein a polybenzimidazole membrane based on polybenzimidazole and phosphoric acid is used in the membrane unit.

    19. The process according to claim 13, wherein the electrochemical separation process in the membrane unit is conducted at a temperature of 120 to 200° C.

    20. The process according to claim 13, wherein the water loading in step (iii) is 0.015 to 0.035 mol of H2O/mol of natural gas.

    21. The process according to claim 13, wherein a gas-gas heat exchanger simultaneously heats up the gas substream and cools down the retentate.

    22. The process according to claim 13, wherein the retentate is dried in a drying stage.

    23. The process according to claim 22, wherein the hydrogen stream obtained from the process of drying the retentate is used for moistening the gas substream in step (iii).

    24. The process according to claim 13, wherein the depletion level of a gas substream is ascertained with the aid of the following formulae:
    m.sub.TG,var,α,opt=(M.sub.PL/M.sub.H2)*(1/AG.sub.var,opt)*(1+1/Y.sub.PL,H2,α)*m.sub.H2,Pr where: M.sub.PL=molar mass of the hydrogen-natural gas mixture in the pipeline, M.sub.H2=molar mass of hydrogen=2.0159 g/mol, AG.sub.var,opt=optimal depletion level of the molar amount of hydrogen in gas substream, calculated by the method that follows, Y.sub.PL,H2,α=molar proportion of hydrogen in the hydrogen-natural gas mixture in the pipeline, m.sub.H2,Pr=desired mass flow rate of pure hydrogen, where: in the following formula for the H2 product-specific variable separation energy expenditure tavar, the value for the depletion level AG is varied by iteration until the value for tavar is at its lowest:
    ta.sub.var=(1/AG)*[K.sub.moist*[(Y.sub.TG,H2O,ω−Y.sub.TG,H2O,α)/Y.sub.PL,H2,α]+K.sub.comp*(1+1/Y.sub.PL,H2,α)*In[(p1+Δp)/p1)])+K.sub.F*{U.sub.EHS,over+K.sub.U,min*In[1+(1+Y.sub.TG,H2O,ω)/Y.sub.PL,H2,α/(1−AG)]} with the following constants: K.sub.moist=13.78 kW.sub.el/kg H2, K.sub.comp=0.49 kW.sub.el/kg, K.sub.U,min=0.019 V, and the process parameters: P1=pipeline pressure, Δp=the total pressure drop between the withdrawal station and the refeeding station for the gas substream from and to the pipeline, U.sub.EHS,over=R.sub.EHS,specific*i.sub.EHS=0.055V with conduction resistances R.sub.EHS,specific and current density i.sub.EHS, p.sub.anode=pressure on anode side, and y.sub.TG,H2,ω=molar proportion of H2 at the exit from the separation unit, Y.sub.TG,H2O,α=molar loading of the natural gas with H.sub.2O in the pipeline, and Y.sub.TG,H2O,ω=molar loading of the natural gas with H.sub.2O before entry to the separation unit.

    Description

    [0061] It was thus an object of the present invention to indicate a process which, with the aid of just a single membrane stage, can separate hydrogen from a pipeline, especially from a pipeline at low positive pressure, and hence indicates inexpensive decentralized provision of high-purity hydrogen.

    [0062] It was another object of the invention to indicate an optimized partial gas volume that is to be separated for pipeline, with which the lowest hydrogen separation costs arise with use of an EHS.

    [0063] What is also being sought is a guideline as to what partial gas volume should be separated from the pipeline in order that the lowest separation costs arise in the separation of hydrogen with the aid of an EHS.

    [0064] Surprisingly, a one-stage membrane process for electrochemical separation of hydrogen from natural gas streams in a pipeline (1) having a positive pressure in the range from 50 mbar to 100 bar has been found, which has the following process steps (FIG. 1): [0065] (i) a gas substream (2) is drawn off from the natural gas stream in a pipeline (1) without any change in the gas composition, where the mass flow rate of the gas substream is adjusted depending on the hydrogen content in the natural gas stream (1) such that a depletion level of 0.65 to 0.975 is established in the case of a hydrogen concentration of <10% by volume and a depletion level of 0.55 to 0.925 in the case of a hydrogen concentration of >10% by weight, where the depletion level is defined as the quotient of the desired molar H2 product stream (6) and the molar H2 reactant flow rate in the gas substream at the inlet of the membrane unit (2), [0066] (ii) this gas substream (2) is compressed (3) upstream of a membrane unit (5), [0067] (iii) this gas substream is heated to 100 to 250° C. either upstream of the membrane unit or in the membrane unit, and this gas substream is supplied with water (4) upstream of the membrane unit and/or on the permeate side of the membrane unit (6a), such that the water loading is between 0.005 and 0.2 mol of water/mol of natural gas, (iv) this gas substream is sent to an electrochemical membrane unit in which hydrogen is separated off as permeate (6a) at a temperature of 100 to 250° C., [0068] (v) the retentate (8) from the membrane unit is recycled into the natural gas stream at the feed point (16), sent to a chemical utilization and/or used as fuel.

    [0069] In the description that follows, all percentages are by volume, unless it is stated explicitly that they are by mass. In the gas phase, mol % corresponds to the percentage by volume.

    Step (i):

    [0070] From a hydrogen-containing natural gas stream being conveyed in a pipeline is taken, for example with the aid of one or more compressors, a gas substream of up to 80% by volume, preferably 0.0001% to 80% by volume, of the total natural gas stream, preferably up to 50% by volume, preferably 0.0001% to 50% by volume, further preferably up to 30% by volume, preferably 0.0001% to 30% by volume, further preferably of 20% by volume, preferably 0.0001% to 20% by volume, especially up to 15% by volume, preferably 0.01% to 15% by volume. The composition of the gas substream (2) corresponds to the composition of the natural gas mixture (1) in the pipeline, i.e. the gas substream (2) is taken without any change in the gas composition. The design of a tapping point for taking of natural gas substreams is known to the person skilled in the art. For instance, the tapping point for taking of a natural gas substream may advantageously be designed, for example, as follows: A volumeter measures the volume flow rate taken (2). An electronic device compares this volume with a target value. If the partial gas flow rate taken is below the target value, the compressor (3) is driven so as to convey a greater volume. If the partial gas flow rate taken is above the target value, the compressor (3) is driven so as to convey a smaller volume. The target value derives from the optimal depletion level AG which is described in detail below.

    [0071] The positive pressure in the pipeline is advantageously 50 mbar to 100 bar, preferably 100 mbar to 20 bar, further preferably 100 mbar to 10 bar, further preferably 100 mbar to 5, further preferably 100 mbar to 4 bar, other preferably 100 mbar to 3 bar, preferably 100 mbar to 2 bar, especially 100 mbar to 1 bar. The pipeline is advantageously a low-pressure and medium-pressure pipeline.

    [0072] Advantageously, the hydrogen-containing natural gas stream comprises 0.1% to 30% by volume of hydrogen, preferably 0.5% to 10% by volume of hydrogen, more preferably 0.5% to 5% by volume of hydrogen, further preferably 0.5% to 3% by volume of hydrogen, especially 1.0% to 2% by volume of hydrogen.

    [0073] The energy expenditure for the separation of the hydrogen from the natural gas stream is determined by the hydrogen content in the natural gas mixture, and also by the necessary water moisture content for the EHS and by the pressure drop over the total distance between the draw point and the point at which the hydrogen-depleted gas substream is fed back to the pipeline. The interplay of these three influences is complex and would lead to incorrect conclusions without a meaningful index. Therefore, the index of depletion level AG is introduced, with the aid of which the lowest apparatus complexity and energy expenditure can be ascertained for any scenario. The optimal mass flow rate is then calculated from the optimal depletion level.

    [0074] The depletion level AG is defined as the ratio of the molar H2 reactant mass flow rate present in the gas substream at the inlet of the membrane unit EHS, n.sub.TG,H2a, and the desired molar H2 product flow rate (6) n.sub.H2,Pr in the formula:


    AG=n.sub.H2,Pr/n.sub.TG,H2,α

    [0075] The person skilled in the art is able to use the specified depletion level to determine the preferred mass flow rate of the gas substream and to adjust it via the output of the compressor(s) in step (ii).


    m.sub.TG,var,α,opt=(M.sub.PL/M.sub.H2)*(1/AG.sub.var,opt)*(1+1/Y.sub.PL,H2,α)*m.sub.H2,Pr  (1)

    [0076] For the establishment of the optimized depletion level, preference is given to using two mass flow rate meters for measurement of the hydrogen content and of the mass flow rate, preferably in the gas substream (2) and in the permeate stream (6a).

    [0077] When the hydrogen loading of the pipeline is <10% by volume, the optimized depletion level is advantageously 0.65 to 0.975, especially 0.85 to 0.975.

    [0078] Preferably, in the case of a hydrogen loading of the pipeline of <5% by volume, an optimized depletion level of 0.75 to 0.975, further preferably 0.8 to 0.975, especially 0.85 to 0.975, is chosen.

    [0079] Preferably, in the case of a hydrogen loading of the pipeline of 5% to <10% by volume, an optimized depletion level of 0.65 to 0.95, further preferably 0.8 to 0.95, especially 0.825 to 0.925, is chosen.

    [0080] When the hydrogen loading of the pipeline is 10% to <20% by volume, the optimized depletion level is advantageously 0.55 to 0.925, further preferably 0.7 to 0.9, especially 0.75 to 0.9. Preferably, in the case of a hydrogen loading of the pipeline of 10% to <15% by volume, an optimized depletion level of 0.6 to 0.925, further preferably 0.7 to 0.925, especially 0.8 to 0.9, is chosen.

    [0081] Preferably, in the case of a hydrogen loading of the pipeline of 15% to >20% by volume, an optimized depletion level of 0.55 to 0.9, further preferably 0.7 to 0.9, especially 0.725 to 0.85, is chosen.

    Step (ii):

    [0082] This gas substream is compressed with an apparatus for increasing the pressure (e.g. fan or compressor) in order either to counterbalance the pressure drops with respect to the pipeline between the withdrawal and refeeding station along the gas substream pathway and/or to build up the necessary pressure for the H2 membrane separation, advantageously 50 mbar to 1 bar. The pressure drop between the withdrawal station and the refeeding station is typically 1 mbar to 5 bar, preferably 10 mbar to 2 bar, especially 50 mbar to 1 bar. This apparatus for increasing the pressure may in principle be positioned anywhere in the gas substream pathway between the withdrawal station and the refeeding station; the determination of the optimized position is known to the person skilled in the art. In the case of low pressure drops, advantageously in the range of 1 mbar to 10 mbar, the apparatus for increasing the pressure is advantageously positioned downstream of the membrane unit and upstream of the refeeding station in flow direction of the gas substream. In the case of high pressure drops, advantageously in the range of 1 bar to 5 bar, the apparatus for increasing the pressure is advantageously positioned between the withdrawal station and upstream of the membrane unit in flow direction of the gas substream.

    [0083] In some cases, for example at the end of a pipeline run, it is advantageous to reduce the pressure of the pipeline stream (12), example with the aid of a throttle or a turbine.

    Step (iii):

    [0084] In step (iii), the gas substream is heated to 100 to 250° C., preferably to 100 to 200° C. Suitable apparatuses for heating the gas substream are known to the person skilled in the art; advantageously, a gas-gas heat exchanger is used. Preference is given to heating up the gas substream (2) prior to the addition of water. Particular preference is given to simultaneously heating up the gas substream (2) and cooling down the retentate (8) in a gas-gas heat exchanger. Preference is given to cooling down the retentate (8) before the drying process (9).

    [0085] Preference is given to adding water to the gas substream for moistening of the gas substream. The manner in which the water is supplied is known to the person skilled in the art. The water content needed differs according to the membrane technology and is known to the person skilled in the art. In the case of EHS within a temperature range of advantageously 120 to 200° C., the preferred water loading is 0.005 to 0.2 mol of H2O/mol of natural gas, further preferably 0.005 to 0.1 mol of H2O/mol of natural gas, further preferably 0.01 to 0.05 mol of H2O/mol of natural gas, further preferably 0.015 to 0.035 mol of H2O/mol of natural gas, especially 0.02 to 0.03 mol of H2O/mol of natural gas. This addition of water can be effected upstream of the membrane unit and/or on the permeate side of the membrane unit; in the case of EHS, the cathode space is the permeate side. Preference is given to the addition of water upstream of the membrane unit.

    Step (iv):

    [0086] In the membrane unit, hydrogen is separated from the gas substream, i.e. hydrogen is formed as permeate. Depending on the membrane unit used, the permeate comprises too much water for the subsequent use. In this case, it is advantageous to conduct a drying process on the permeate (7), for example temperature swing adsorption (TSA), drying with adsorber materials such as silica gel, CaO, superabsorbents or by glycol scrubs. The operating conditions of the TSA are known to the person skilled in the art and are described, for example, in https://www.reicat.de/de/gastrocknung.html.

    [0087] Advantageous applications for the use of the hydrogen separated are hydrogen filling stations and decentralized hydrogen users. Moreover, this process offers the option of a hydrogen-free natural gas stream for processes that are sensitive to hydrogen, for example natural gas caverns, chemical processes, turbines.

    Membrane Unit:

    [0088] A preferred membrane unit is the EHS, which uses a phosphoric acid-doped membrane. The setup for hydrogen separation is largely identical to a fuel cell setup. The core of the EHS system is the membrane electrode assembly (MEA). On the anode side, hydrogen molecules are oxidized on a catalyst to protons, which pass through the proton-selective membrane to the cathode side, while electrons migrate through an external electrical circuit to the cathode. As long as current is being applied, the EHS system thus separates hydrogen from gas mixtures. The technology of electrochemical hydrogen separation is described for example in WO 2016/50500 and WO 2010/115786.

    [0089] The catalytically active material used may be the customary compounds and elements known to those skilled in the art that catalyze the dissociation of molecular hydrogen into atomic hydrogen, the oxidation of hydrogen to protons, and the reduction of protons to hydrogen. Suitable examples are Pd, Pt, Cu, Ni, Ru, Fe, Co, Cr, Mn, V, W, tungsten carbide, Mo, molybdenum carbide, Zr, Rh, Ru, Ag, Ir, Au, Re, Y, Nb, and alloys and mixture thereof, with preference in accordance with the invention given to Pt. The catalytically active materials may also be present in supported form, preferably with carbon as support. In a further development of the membrane electrode assembly, the amount of the catalytically active material in the cathode catalyst is 0.1 mg/cm2 to 2.00 mg/cm2, preferably 0.1 mg/cm2 to 1 mg/cm2, based on the total surface area of the anode and cathode.

    [0090] The membrane used in accordance with the invention selectively conducts protons, that is to say, in particular, that it is not electron-conducting. In accordance with the invention, it is possible to use for the membranes all materials known to those skilled in the art from which proton-conducting membranes can be formed. It is also possible to use in accordance with the invention selectively proton-conducting membranes such as are known from fuel-cell technology.

    [0091] Materials that are particularly suitable for the production of gas-tight and selectively proton-conducting membranes are polymer membranes. Suitable polymers are sulfonated polyether ether ketones (S-PEEK), sulfonated polybenzimidazoles (S-PBI), and sulfonated fluorinated hydrocarbon polymers (for example Nafion®). It is also possible to use perforated polysulfonic acids, styrene-based polymers, poly(arylene ethers), polyimides, and polyphosphazenes.

    [0092] Very particular preference is given to using membranes made of polybenzimidazoles, especially MEAs based on polybenzimidazole and phosphoric acid, such as those marketed under the Celtec-P® name by BASF SE.

    [0093] The operating conditions of the EHS system are strongly dependent on the MEA chosen. When using the Celtec® technology, the use of a voltage of 0.1 to 0.4 V and a current of 0.2 to 1 A/cm.sup.2 is advantageous. The separation of H2 is based not on differential pressure, but on electrochemistry. EHS can therefore be operated advantageously at ambient pressure. Provided there is no differential pressure between the anode and the cathode, a higher pressure is advantageous, which results in a higher separation rate.

    [0094] The hydrogen separation rate is typically between 60% and 99%, preferably 70 to 95%, especially 80% to 90%, wherein the higher the separation rate, the higher the electrical energy requirement of an EHS.

    [0095] The current density is advantageously 0.1 to 1 A/cm.sup.2, preferably 0.2 to 0.7 A/cm.sup.2, especially 0.2 to 0.5 A/cm.sup.2. The voltage is advantageously 1 to 1000 mV, preferably 100 to 800 mV, especially 150 to 350 mV.

    [0096] These electrochemical hydrogen separation systems are operated at temperatures of advantageously 120 to 200° C., preferably 150 to 180° C., especially 160 to 175° C. The pressure is advantageously 0.5 to 40 bar, preferably 1 to 10 bar, especially 1 to 5 bar. The pressure difference between the anode side and the cathode side is advantageously less than 1 bar, preferably less than 0.5 bar.

    [0097] This mode of operation allows a high tolerance to gas impurities, for example CO (3%) and H2S (15 ppm), to be achieved.

    [0098] This relatively low temperature permits relatively rapid and material-sparing start-up and shutdown, which is an advantage especially for non-continuous operation in decentralized systems with fluctuating hydrogen output, for example in filling stations.

    [0099] The active surface area of the membrane electrode assembly is advantageously within a range from 5 to 20 000 cm.sup.2, preferably 25 to 10 000 cm.sup.2, especially 150 to 1000 cm.sup.2.

    [0100] The thickness of the membrane electrode assembly is advantageously within a range from 250 to 1500 μm, preferably 600 to 1000 μm.

    [0101] At a construction volume of 1 m.sup.3, a hydrogen separation stack consisting of end plates, bipolar plates, seals, and membrane electrode assemblies advantageously separates 100 to 200 Nm.sup.3/h hydrogen and is accordingly significantly smaller than systems with physical hydrogen separation.

    [0102] The energy consumption is typically between 3 and 7 kWh/kg H2, depending on the gas composition and chosen separation rate.

    [0103] Since the electrochemical separation is based on gas-tight, highly selective proton-conducting membranes, the purity of the hydrogen generated can be very high, typically greater than around 99.9%, preferably greater than 99.95%, in particular greater than 99.99%.

    [0104] Particular preference is given to the following membrane electrode assembly specifications:

    TABLE-US-00002 Width Thickness Acid concentration Acid content Acid content PBI content I.V. value Specification (mm) (μm) (wt %) (mg/cm.sup.3) (mg/cm.sup.2) (mg/cm.sup.3) (dL/g) Values 310 360-440 51-59 760-870 30-37 65-89 4.50-6.00 PBI = polybenzimidazole I.V. = inherent viscosity

    Depletion Level:

    [0105] The depletion level AG is defined as the ratio of the molar H2 reactant mass flow rate present in the gas substream at the inlet of the membrane unit EHS, n.sub.TG,H2a, and the desired molar H2 product flow rate (6) n.sub.H2,Pr in the form:


    AG=n.sub.H2,Pr/n.sub.TG,H2,α

    [0106] The optimal depletion level with regard to costs is dependent on the required separation energy expenditure for the respective depletion.

    [0107] The depletion level AG leads to the molar partial gas volume flow rate n.sub.TG which is drawn off from the pipeline for separation of the amount of H2 product, n.sub.H2,pr, according to the following formula:


    n.sub.TG=(1/AG)*(1+1/Y.sub.PL,H2,α)*n.sub.H2,Pr

    [0108] The loading Y.sub.PL,H2,α is proportional to the H2 concentration in the natural gas pipeline and is defined as the molar loading of the natural gas n.sub.PL;EG with hydrogen n.sub.PL,H2 before the withdrawal of the gas substream n.sub.TG.


    YY.sub.PL,H2,α=n.sub.PL,H2/n.sub.PL;EG

    [0109] The depletion level is advantageously in the range between 0.5 and 0.999, preferably 0.7 and 0.99, further preferably between 0.80 and 0.99.

    [0110] The term “loading” is known to the person skilled in the art, inter alia, from the description of the composition of moist air (see https://www.thm.de/wi/images/user/Thielen-72/Downloads/Energietechnik/Kapitel_8_−feuchte_Luft.pdf, pages 146 and 147). In the case of moist gases, the composition is typically based on the mass or moles of the dry gas, since these parameters of the dry gas remain constant in all physical processes.

    [0111] The loading YY.sub.PL,H2,α can be converted to % by volume for a two-substance mixture consisting of H2 and CH4 by the following formula:


    yY.sub.PL,H2,α in % by volume=n.sub.PL,H2/(n.sub.PL;EG+n.sub.PL,H2)=YY.sub.PL,H2,α/(1+YY.sub.PL,H2,α)

    [0112] The optimal AG range for a given H.sub.2 loading in the pipeline Y.sub.H2;Pl,α and the given pressure p.sub.1 at the tapping point in the pipeline for the gas substream and a pressure drop Δp that arises from the apparatus situation over the entire pathway of the gas substream and the established loading of the dry gas substream with water vapor Y.sub.TG,H2O,ω results from the requirement that, within this range, the specific energy consumptions ta.sub.var must not be greater than 30% of the minimum specific energy consumption ta.sub.var,min. The requirement of 30% corresponds to the typical approach that the competitor should have production costs at least 30% higher in the operation of an analogous process.

    [0113] For example, with p.sub.1=1.5 bar and Δp=0.1 bar and Y.sub.TG,H2O,ω=0.025. the following optimal AG ranges are found as a function of the H.sub.2 loading in the pipeline Y.sub.H2,Pl,α:

    Y.sub.H2,Pl,α=0.05: AG.sub.opt=0.57 to 0.99996; 0.64 to 0.9997; 0.73 to 0.997
    Y.sub.H2,Pl,α=0.10: AG.sub.opt=0.48 to 0.9993; 0.55 to 997; 0.65 to 0.987
    Y.sub.H2,Pl,α=0.20: AG.sub.opt=0.39 to 0.9958; 0.46 to 0.988; 0.55 to 0.965

    [0114] For example, with Y.sub.H2,Pl,α=0.10 and p.sub.1=1.5 bar and Y.sub.TG,H2O,ω=0.025, the following optimal AG ranges are found as a function of the pressure drop Δp over the entire pathway of the gas substream:

    Δp=0.1 bar: AG.sub.opt=0.48 to 0.9993; 0.55 to 0.997; 0.65 to 0.987
    Δp=1.0 bar: AG.sub.opt=0.56 to 0.99992; 0.63 to 0.9994; 0.73 to 0.996
    Δp=10.0 bar: AG.sub.opt=0.66 to 0.99999; 0.72 to 0.99998; 0.81 to 0.9997

    [0115] For example, with Y.sub.H2,Pl,α=0.1 and p.sub.1=1.5 bar and Δp=0.1 bar, the following optimal AG ranges are found as a function of the H.sub.2O loading of the gas substream Y.sub.TG,H2O,ω:

    Y.sub.TG,H2O,ω=0.01: AG.sub.opt=0.35 to 0.995; 0.41 to 0.986; 0.50 to 0.959
    Y.sub.TG,H2O,ω=0.10: AG.sub.opt=0.64 to 0.99999; 0.71 to 0.99998; 0.80 to 0.9995
    Y.sub.TG,H2O,ω=0.20: AG.sub.opt=0.69 to 0.99999; 0.76 to 0.99998; 0.84 to 0.99997

    Step (v):

    [0116] The retentate from the membrane unit (8), an H2-depleted stream, is recycled into the natural gas stream (1). If the water content of the retentate is too high for this recycling into the pipeline, the retentate is advantageously dried before being introduced into the pipeline. Drying processes (9) are known to the person skilled in the art; advantageous examples are gas scrubbings with ethylene glycols. Drying affords an aqueous stream (11) that can be worked up, for example, by distillation or with the aid of membrane methods. The water stream (11) obtained from the process of drying the retentate can advantageously be used for moistening of the gas substream (4) in step (iii).

    [0117] Alternatively, the retentate from the membrane unit (8) is sent to a chemical utilization and/or used as fuel.

    [0118] Many chemical processes and internal combustion engines can be operated solely with largely pure natural gas. The process for producing benzene from natural gas, for example, requires a separation of hydrogen from the reaction mixture to avoid establishment of a chemical equilibrium and stopping of the reaction. Here, the hydrogen concentration must be reduced from 11% to less than 1%. A further example is natural gas filling stations that supply cars and buses with natural gas, which may comprise only less than 2% hydrogen. Given an addition limit of 10% hydrogen as permitted nowadays, therefore, removal of the hydrogen is absolutely necessary. If appropriate, the retentate is cooled down before being introduced into the pipeline, advantageously from 100 to 260° C. in the membrane unit down to a dew point of −8° C. at pipeline pressure. Suitable apparatuses for cooling the gas substream are known to the person skilled in the art; advantageously, a gas-gas heat exchanger is used. Advantageously, such a gas-gas heat exchanger is used for simultaneous heating of the gas substream in step (iii) and for cooling the gas substream in step (v).

    Method of Ascertaining the Optimal Partial Gas Volume Flow Rate

    [0119] The invention further relates to a method of ascertaining the optimal gas substream which is drawn off from a pipeline that conducts natural gas and hydrogen in order to separate hydrogen from this gas substream in an electrochemical membrane unit. The optimal gas substream incurs the lowest separation costs.

    [0120] Since only a mass flow rate (mass/h) is measurable, the optimal gas substream is determined as a mass flow rate. A molar flow rate (mol/h), by contrast, is of better suitability for the description of molar processes, for example the transport of H2 through the EHS or the moistening of gas streams. Conversion from moles to mass via the molar mass is common practice to the person skilled in the art.

    [0121] Calculation of the optimal gas substream m.sub.TG,opt requires the optimal depletion level AG.sub.opt. This is ascertained by iteration as follows: In the following formula for the H2 product-specific variable separation energy expenditure ta.sub.var, the value for the depletion level AG is varied until the value for ta.sub.var is at its lowest.


    ta.sub.var=(1/AG)*{K.sub.moist*[(Y.sub.TG,H2O,ω−Y.sub.TG,H2O,α)/Y.sub.PL,H2,α]+K.sub.comp*(1+1/Y.sub.PL,H2,α)*In[(p1+Δp)/p1)]}+K.sub.F*{U.sub.EHS,over+K.sub.U,min*In[1+(1+Y.sub.TG,H2O,ω)/Y.sub.PL,H2,α/(1−AG)]}


    with the following constants:


    K.sub.moist=13.78 kW.sub.elkg H2,K.sub.comp=0.49 kW.sub.el/kg,KF=26.59 Ah/kgH2,K.sub.U,min=0.019 V

    and the process parameters:
    U.sub.EHS,min=K.sub.U,min*In (p.sub.cathode/(p.sub.anode*y.sub.TG,H2,ω), with p.sub.cathode=pressure on cathode side, p.sub.anode=pressure on anode side and y.sub.TG,H2,ω=molar proportion of H2 at the exit from the separation unit (8).

    [0122] Δp is the total pressure drop between the withdrawal station (15) and the refeeding station (16) for the gas substream from and to the pipeline.

    [0123] Y.sub.TG,H2O,ω is the molar loading of the natural gas with H.sub.2O in the pipeline (1)

    [0124] Y.sub.TG,H2O,ω is the molar loading of the natural gas with H.sub.2O before entry to the separation unit

    [0125] Y.sub.PL,H2,α=molar loading of the natural gas with H2 in the pipeline (1)

    [0126] The procedure for determination of the optimal depletion level AG.sub.opt may advantageously be as follows:

    [0127] The procedure starts with an AG=0.90, which gives a value for the specific energy consumption ta.sub.var.

    [0128] Subsequently, the value for AG is increased by 0.01 to 0.91, and the corresponding specific energy consumption ta.sub.var is obtained. If the new value for ta.sub.var (AG=0.91) is less than the value before (AG=0.9), the new AG value is increased again by 0.01 in the next step. This procedure is repeated until the new value for ta.sub.var (AG.sub.alt+0.01) is greater than the old (AG.sub.alt). The old AG value then constitutes the optimal depletion level AG.sub.var,opt.

    [0129] If the AG value of 0.91 should lead to a higher value for ta.sub.var than the AG value of 0.90, in the next step, the value for AG is increased by 0.01 each time until the new value for ta.sub.var (AG.sub.alt-0.01) rises again. The old AG value then constitutes the optimal depletion level AG.sub.var,opt with the minimum specific energy consumption ta.sub.var,min.

    [0130] Since the characteristic value per capital costs, expressed as Σ(q.sub.i+p.sub.i), tends to follow the progression for ta.sub.var, the value from the variation of AG is assumed to be the optimal value for ta.sub.var.

    [0131] The economic viability of a process is apparent from the difference between the price of a product and the product costs. The product costs can be calculated from the manufacturing costs and a return on investment (ROI), which reflects the return requirements, interest on capital and the tax situation. If the product costs are above the product price, the process is uneconomic. The production costs are composed of a variable component and a fixed component. In our case, the variable component is dominated by the energy costs. The energy costs here are the product of the cost of power and ta.sub.var.

    [0132] The fixed component, in a first approximation, is dependent on the capital costs. The capital costs are in turn calculated from the product of S(qi+pi) and the specific capital costs that relate thereto.

    [0133] The ROI is directly proportional to the capital costs.

    [0134] qi here is the total heat transfer output and pi is the total electrical power, based in each case on the H2 product mass flow rate mH2,Pr.

    [0135] If the lowest value for ta.sub.var has been ascertained by variation/isolation of AG, the corresponding depletion level AG corresponds to the optimal depletion level AG.sub.opt. With the aid of AG.sub.opt, the optimal gas substream m.sub.TG,var,α,opt is calculated by the formula:


    m.sub.TG,var,α,opt=(M.sub.PL/M.sub.H2)*(1/AG.sub.var,opt)*(1+1/Y.sub.PL,H2,α)*m.sub.H2,Pr

    with
    M.sub.PL=molar mass of the hydrogen-natural gas mixture in the pipeline (1)
    M.sub.H2=molar mass of hydrogen=2.0159 g/mol
    m.sub.H2,Pr=desired mass flow rate of pure hydrogen (6a)

    Advantages:

    [0136] The advantages of the present process are as follows: (i) one-stage process, (ii) efficient process, especially for low hydrogen concentrations, e.g. <5 bar, (iii) process tolerant to impurities in the feed gas stream, (iii) electrochemical separation in operation at ambient pressure, (iv) low space requirement, (v) modular principle: extension of capacity by installation of a further stack, (vi) advantageous for relatively small amounts of gas.

    [0137] The advantage in the use of the method is that, for the wide variety of different hydrogen contents in the pipeline and for the wide variety of different pipeline pressures and for the wide variety of different electrochemical hydrogen separation technologies for a desired product stream, the optimal gas substream is not always at a depletion level of 1, but varies and can thus be ascertained. The optimal gas substream incurs the lowest separation costs.

    FIG. 1 Shows the Process of the Invention for the Case that the Product is High-Purity H2.

    [0138] A gas substream 2 is drawn off from a H2-containing natural gas stream 1 conveyed in a pipeline at a withdrawal point 15 without any change in the gas composition. This gas substream is compressed with a machine for increasing pressure 3 (e.g. fan or compressor) in order either to compensate solely for the pressure drops along the gas substream pathway or in order to build up the necessary pressure for separation of H2. This machine for increasing pressure may in principle be anywhere in the gas substream pathway. The person skilled in the art in the field of chemical engineering will know the best position.

    [0139] Water 4 is added to the gas substream 2 for moistening of the gas substream 2. The water content needed is different according to the membrane technology It is also conceivable that the membrane is also supplied or solely supplied on the permeate side with water 4a to prevent it from drying out. In case of the EHS, the cathode space is the permeate side.

    [0140] In the H2 separation unit 5, H2 is separated from the gas substream 2a. The membrane unit 5 is an electrochemical hydrogen separation, EHS. The driving force for the transport of H2 here is the natural logarithm of the electrical potential ratio.

    [0141] The permeate is the hydrogen 6a. According to the technology, the permeate may still be too moist for the subsequent use. Then the stream 6a is advantageously subjected to a drying process 7, for example temperature swing adsorption. The stream 6 is then on-spec hydrogen. The retentate 8 in the case of an EHS is the anode off-stream.

    [0142] If this H2-depleted stream is too moist for recycling into the pipeline at the refeeding point 16, it is advantageously subjected to drying 9 prior to introduction. Examples of useful drying processes include gas scrubbings with ethylene glycols. This affords an aqueous stream that can be worked up by distillation or with the aid of membrane methods. The water stream 11 obtained from the drying process could in principle be recirculated again to the moistening of the gas substream 2.

    [0143] The dried low-H2 gas stream 10 is then fed back to the residual gas stream 12 in the pipeline grid.

    [0144] In particular cases, it may be more favorable to dispense with the machine for increasing pressure 3 in the gas substream 2 and instead to install an apparatus for generating a pressure drop 14 (e.g. a throttle or a turbine) into the gas stream 12.

    FIG. 2 Shows the Process of the Invention for the Case that the Product is H2-Free Natural Gas.

    [0145] There are natural gas-consuming processes that require a specification with regard to the hydrogen content in the natural gas (FIG. 2). In this case, the gas substream 2 drawn off is likewise compressed (3) and moistened (4) if required before being fed to the membrane unit (5).

    [0146] Subsequently, the retentate stream (8) may still be dried (9) if required. The dried retentate stream (10) is then fed not into the pipeline but to the natural gas-consuming process (17).

    FIG. 3 Shows the Process Example.

    [0147] 40 kg/h of H2 in ultrahigh purity (stream 7) are first to be separated from a pipeline in which 29 050 kg/h of natural gas (stream 1) is being conveyed at a positive pressure of 500 mbar, which initially comprises 10% by volume of H2. Upstream of the site of introduction of the H2 into the natural gas pipeline grid, the natural gas comprises 0.2% by volume of water. This corresponds to a dew point of −8° C. at 1.5 bar.sub.absolute. The H2 is to be separated with the aid of an EHS which is operated at ambient pressure and at 160° C. and which requires a water content in gas stream 6 of 2.5% by volume upstream of the EHS for reliable operation thereof.

    [0148] In simplified terms, natural gas is to be considered as pure methane. In addition, the index PL hereinafter represents pipeline, the index a represents the start and the index ω represents the end. The indices H2, H.sub.2O and CH4 represent the respective component.

    [0149] FIG. 3 shows the influence of the H2 loading of the natural gas in the pipeline YH2,PL, α on the specific energy consumption ta.sub.var. The separation unit here moistens the gas substream to 0.025 mol H2O/mol of natural gas, and the total pressure drop over the gas substream pathway is 0.1 bar.

    [0150] It is apparent that the optimal depletion level AG is in the range between 0.80 and 0.95. The optimal depletion level AG.sub.opt is where the specific energy consumption ta.sub.var has its lowest value. It is also apparent that the higher the H2 loading of the natural gas in the pipeline YH2,Pl, α, the lower the optimal depletion level AG.sub.opt. In the case that YH2,Pl, α=0.05, AG.sub.opt is 0.95, and in the case that YH2,Pl, α=0.20, AG.sub.opt is 0.80.

    [0151] FIG. 4 shows the influence of the necessary moistening of the gas substream on the specific energy consumption ta.sub.var for different H.sub.2O loadings of the natural gas YH2O;TG, ω in the gas substream upstream of the separation unit. The natural gas, in its original state, has an H.sub.2O loading of YTG,H2O, α=YPL,H2O, α=0.00223 mol of H2O/mol of natural gas. The H2 loading of the natural gas in the pipeline YH2,PL, α in this example is 0.10 mol H2/mol of natural gas, and the total pressure drop over the gas substream pathway is 0.1 bar.

    [0152] It is apparent that the optimal depletion level AG is in the range between 0.70 and 0.97. It is also apparent that the lower the necessary moistening, the lower the optimal depletion level AG.sub.opt. In the case that YTG,H2O, ω=0.010, AG.sub.opt is 0.70, and in the case that YTG,H2O, ω=0.200, AG.sub.opt is 0.97.

    [0153] FIG. 5 shows the influence of the pressure drop over the entire pathway of the gas substream from the withdrawal station to the refeeding station into the pipeline on the specific energy consumption ta.sub.var. The H2 loading of the natural gas in the pipeline YH2,PL, α in this example is 0.10 mol of H2/mol of natural gas, and the pipeline pressure p1 is 1.5 bar.sub.abs. Upstream of the separation unit, the gas substream is moistened to 0.025 mol of H2O/mol of natural gas.

    [0154] It is apparent that the optimal depletion level AG is in the range between 0.85 and 0.97. It is also apparent that the lower the pressure drop, the lower the optimal depletion level AG.sub.opt. In the case that Δp=0.1 bar, AG.sub.opt is 0.85, and in the case that Δp=10.0 bar, AG.sub.opt is 0.97.

    [0155] FIG. 6 shows the influence of the H2 loading of the natural gas in the pipeline YH2,PL, α on the separation energy expenditure Σ(qi+pi). The capital costs for the overall plant are proportional to the separation energy expenditure.

    [0156] The separation unit in this example moistens the gas substream to 0.025 mol H2O/mol of natural gas, and the total pressure drop over the gas substream pathway is 0.1 bar.

    [0157] It is apparent that the optimal depletion level AG is in the range between 0.96 and 0.99. The optimal depletion level AG.sub.opt is where the separation energy expenditure Σ(qi+pi) has its lowest value. It is also apparent that the higher the H2 loading of the natural gas in the pipeline YH2,PI, α, the lower the optimal depletion level AG.sub.opt. In the case that YH2,PI, α=0.05, AG.sub.opt is 0.99, and in the case that YH2,PI, α=0.20, AG.sub.opt is 0.96.

    [0158] FIG. 7 shows the influence of the necessary moistening of the gas substream on the separation energy expenditure Σ(qi+pi) for different H.sub.2O loadings of the natural gas YH2O;TG, ω in the gas substream upstream of the separation unit. The natural gas, in its original state, has an H2O loading of YTG,H2O, α=YPL,H2O, α=0.00223 mol of H2O/mol of natural gas. The H2 loading of the natural gas in the pipeline YH2,PL, α in this example is 0.10 mol H2/mol of natural gas, and the total pressure drop over the gas substream pathway is 0.1 bar.

    [0159] It is apparent that the optimal depletion level AG is in the range between 0.90 and 0.97. It is also apparent that the lower the necessary moistening, the greater the optimal depletion level AG.sub.opt. In the case that YTG,H2O, ω=0.010, AG.sub.opt is 0.97, and in the case that YTG,H2O, ω=0.200, AG.sub.opt is 0.90.

    [0160] FIG. 8 shows the influence of the pressure drop over the entire pathway of the gas substream from the withdrawal station 15 to the refeeding point 16 into the pipeline on the separation energy expenditure Σ(qi+pi).

    [0161] The H2 loading of the natural gas in the pipeline YH2,PL, α in this example is 0.10 mol of H2/mol of natural gas, and the pipeline pressure is 1.5 bar.sub.abs. Upstream of the separation unit, the gas substream is moistened to 0.025 mol of H2O/mol of natural gas.

    [0162] It is apparent that the optimal depletion level AG is in the range between 0.97 and 0.99. It is also apparent that the lower the pressure drop, the lower the optimal depletion level AG.sub.opt. In the case that Δp=0.1 bar, AG.sub.opt is 0.97, and in the case that Δp=10.0 bar, AG.sub.opt is 0.99.

    1) Pipeline

    [0163] For a gas mixture stream 1 with 10% by volume of hydrogen in the pipeline, an average molar mass M.sub.PL of 14.64 kg/kmol is calculated. This leads to a value for the molar natural gas flow rate n.sub.PL,α:


    n.sub.PL,α=m.sub.PL,α/M.sub.PL=29 050 kg/h/14.64 kg/kmol=1984.2 kmol/h  (1.1)

    [0164] The molar amount of H2 present in the gas mixture stream 1 upstream of the withdrawal point, n.sub.PL,H2,α, is


    n.sub.PL,H2,α=y.sub.H2,α*n.sub.PL,α=1984.2*0.10=198.4 kmol/h  (1.2)

    [0165] The molar amount of H.sub.2O present in the gas mixture stream 1 upstream of the withdrawal point, n.sub.PL,H2O,α, is


    n.sub.PL,H2O,α=Y.sub.H2O,α*n.sub.PL,α=1984.2*0.002=3.97 kmol/h  (1.3)

    [0166] The molar amount of CH4 present in the gas mixture stream 1, nPL,CH4, a, is


    n.sub.PL,CH4,α=n.sub.PL,α−n.sub.PL,H2,α−n.sub.PL,H2O,α=1984.2-198.4-3.97=1781.83 kmol/h  (1.4)

    [0167] This leads to the following loadings Y based on the amount of CH4:


    Y.sub.PL,H2,α=n.sub.PL,H2,α/n.sub.PL,CH4,α=198.4/1781.83=0.111  (1.5)


    Y.sub.PL,H2O,α=n.sub.PL,H2O,α/n.sub.PL,CH4,α=3.97/1781.83=0.00223  (1.6)

    [0168] The H2 product flow rate n.sub.H2,Pr desired for operation of a filling station is


    n.sub.H2,Pr=m.sub.H2,Pr/M.sub.H2=40 kg/h/2 kg/kmol=20 kmol/h  (1.7)

    2) Gas Substream

    [0169] The H2 is not separated directly from the gas mixture stream in the pipeline (1), but from a significantly smaller gas substream (2).

    [0170] If a depletion level AG is defined as follows:


    AG=(n.sub.TG,H2,α−n.sub.TG,H2,ω)/n.sub.TG,H2,α=n.sub.H2,Pr/n.sub.TG,H2,α=0.990  (2.1)

    with the amount of H2 present in gas substream 2 at the inlet to the EHS, n.sub.TG,H2,α, and the amount remaining in gas substream 8 at the exit from the EHS, n.sub.TG,H2,ω, this is directly correlated to the total partial gas volume present in gas substream 2 that has been drawn off from the pipeline stream, n.sub.TG,α:


    n.sub.TG,α=(1/AG)*(1+1/Y.sub.PL,H2,α)*n.sub.H2,Pr=200 kmol/h  (2.2)

    [0171] All loadings hereinafter are based on the amount of CH or present in gas substream 2, since this remains constant over all process steps. This is because:

    [0172] With the loading


    Y.sub.TG,H2,α=Y.sub.PL,H2,α=n.sub.TG,H2,α/n.sub.TG,CH4,α  (2.3)

    and equation (2.1), it follows that


    n.sub.TG,CH4,α=n.sub.TG,CH4,ω=n.sub.TG,CH4=n.sub.TG,H2,α/Y.sub.PL,H2,α=n.sub.H2,Pr/AG/Y.sub.PL,H2,α=180 kmol/h  (2.4)

    with the ratio


    X=n.sub.H2,Pr/n.sub.PL,H2,α=0.100  (2.5)

    3) Increase in Pressure

    [0173] In order to overcome flow losses in the gas substream from 2 to 12, this must be compressed. In the ideal case, this is done isothermally. In this case, the compressor output P1 needed for this purpose in order to achieve an illustrative pressure increase of Δp=100 mbar is as follows:


    P1=m.sub.TG,α*(R*T.sub.PL/M.sub.PL/η.sub.comp)*In((p.sub.1+Δp)/p.sub.1)=12 kW  (3.1)

    [0174] For the gas substream m.sub.TG,a, after some rearrangements, the following relationship is found:


    m.sub.TG,α=m.sub.H2,Pr*(K.sub.M,TG,α/AG)*(1+1/Y.sub.PL,H2,α)=2934 kg/h  (3.2)


    with


    K.sub.M,TG,α=M.sub.TG,α/M.sub.H2=M.sub.PLM.sub.H2=7.26  (3.3)

    [0175] For the specific compressor output p.sub.comp, it thus follows that:


    p.sub.comp=P1/m.sub.H2,Pr=K.sub.comp*(1/AG)*(1+1/Y.sub.PL,H2,α)*In((p.sub.1+Δp)/p.sub.1)=0.30 kW.sub.el/kgH2  (3.4)


    with


    K.sub.comp=R*T.sub.PL/M.sub.H2/η.sub.comp=0.49 kW.sub.el/kgH2  (3.5)

    [0176] And a derived compressor efficiency η.sub.comp of


    η.sub.comp=0.70  (3.6)

    4) Heating/Cooling of Gas Substream

    [0177] The isothermally compressed gas substream 3 has to be heated upstream of the EHS from the pipeline temperature τ.sub.PL=25° C. to that temperature of τ.sub.EHS=160° C. and then cooled down again. For this purpose, the heat flow Q1 has to be transferred in a heat exchanger.

    [0178] With the approximate value for cp.sub.TG


    cp.sub.TG=2.3 kJ/kgK  (4.1)


    it follows that


    Q1=m.sub.TG,α*cp.sub.TG*(τ.sub.EHS−τ.sub.PL)=253 kW  (4.2)

    [0179] The gas substream, before being introduced into the natural gas grid, has to be dried again to the water content of the natural gas pipeline of 0.2 mol %. In the case of drying with propylene glycol, for this purpose, according to a thermodynamic simulation for 1.5 bar.sub.absolute, the gas substream has to be cooled down to T.sub.drying=21° C. For this purpose, the heat flow Q2 has to be transferred in a heat exchanger.


    Q2=m.sub.TG,α*cp.sub.TG(τ.sub.PL−τ.sub.drying)=8 kW  (4.3)


    With


    K.sub.HT,TG=cp.sub.TG*((τ.sub.EHS−τ.sub.drying)=0.089 kWh/kg  (4.4)

    it follows that, for the heat flow to be transferred based on the product mass flow rate m.sub.H2,Pr, or the heat transfer output for the gas substream,


    q.sub.TG=(Q1+Q2)/m.sub.H2,Pr=K.sub.HT,TG*(K.sub.M,TG,α/AG)*(1+1/Y.sub.PL,H2,α=6.50 kWh/kgH.sub.2  (4.5)

    5) Separation of H2 with EHS

    [0180] According to Faraday's law, for the electrochemical separation of stream 7 as H2 product volume flow rate n.sub.H2,Pr=19.84 kmol/h, the following electrical current I.sub.EHS is needed:


    I.sub.EHS=2*n.sub.H2,pr*F=1 063 600 A  (5.1)


    with F=96485 AS/mol  (5.2)

    The minimum voltage necessary for separation, U.sub.EHS,min, is calculated by the following formula:


    U.sub.EHS,min=K.sub.U,min*In(p.sub.cathode/(p.sub.anode*y.sub.TG,H2,ω)=0.127 V  (5.3)


    with


    K.sub.U,min=R*T.sub.EHS/2/F=0.019 V  (5.4)


    with the ideal gas constant R


    R=8.314 J/mol K  (5.5)

    and the partial H2 pressure in the cathode space p.sub.H2,cathode=1.013 bar


    p.sub.H2,cathode=p.sub.cathode*y.sub.H2,cathode=1.013 bar  (5.6)


    with y.sub.H2,cathode=1


    and


    p.sub.anode=1.013 bar  (5.7)

    [0181] For the necessary separation energy, the partial pressure at the exit from the EHS, p.sub.H2,anode, is crucial, i.e.


    p.sub.H2,anode=p.sub.anode*Y.sub.TG,H2ω=0.0011  (5.8)

    [0182] With equation (6.2) for Y.sub.TG,H2,ω, it follows that:


    y.sub.TG,H2,ω)=1/(1+(1+Y.sub.TG,H2Oω)/Y.sub.PL,H2α/(1−AG))=0.0011  (5.9)

    [0183] In the EHS, however, conduction resistances R.sub.EHS,specific also occur across the membrane. These are dependent on the current density i.sub.EHS and the CO content of the gas mixture. The current density i.sub.EHS is normally in the order of magnitude of 0.1 to 1 A/cm.sup.2.

    [0184] For a CO-free gas mixture and an assumed current density i.sub.EHS of


    i.sub.EHS=0.8 A/cm.sup.2  (5.10)

    the specific membrane resistance R.sub.EHS,specific is calculated for the CHS, for example, by the following formula:


    R.sub.EHS,specific=98-20*(i.sub.EHS.sup.2+i.sub.EHS)=69.2 mΩ*cm.sup.2  (5.11)

    [0185] The overvoltage component for the CHS, U.sub.EHS,over is thus calculated as follows:


    U.sub.EHS,over=R.sub.EHS,specific*i.sub.EHS=0.055 V  (5.12)

    [0186] This leads to the total voltage necessary for the separation of H2, U.sub.EHS,tot:


    U.sub.EHS,tot=U.sub.EHS,min+U.sub.EHS,over=0.182 V  (5.13)

    [0187] The necessary electrical power P2 for the EHS is then found from


    P2=U.sub.EHS,tot*I.sub.EHS=194 kW  (5.14)

    or the electrical power p.sub.EHS based on the H2 product mass flow rate


    p.sub.EHS=P.sub.EHS/m.sub.H2,Pr=4.84 kWh/kg H2  (5.15)

    6) Moistening/Drying

    [0188] The water content y.sub.TG,H20,ω) in gas substream 6 upstream of the EHS should, by way of example, be 2.5% by volume in order to prevent the membrane from drying out. In order to attain that value, a water vapor stream Δn.sub.TG,H2O is added via stream 5 to the partial gas volume flow 6.

    [0189] With equation (1.6) and


    Y.sub.TG,H2O,α=Y.sub.PL,H2O,α=0.00223  (6.1)


    and


    Y.sub.TG,H2,ω=(1+Y.sub.PL,H2,α/(1/y.sub.TG,H2O,ω−1)=0.0285  (6.2)


    it follows with equation (2.4) that


    Δn.sub.TG,H2O=n.sub.TG,CH4*(Y.sub.TG;H2O,ω−Y.sub.TG,H2O,α)=4.73 kmol H2O/h  (6.3)


    or


    Δm.sub.TG,H2O=Δ.sub.nTG,H2O*M.sub.;H2O=85.1 kg H2O/h  (6.4)

    [0190] This additionally introduced amount of water vapor Δm.sub.TG,H2O, downstream of the EHS, must be separated again from gas substream 8 in order to prevent condensation after the H2-depleted gas substream 12 has been fed back to the pipeline.

    [0191] According to the prior art, the gas substream 9 is dried with the aid of a glycol scrub. This requires both steam and power for operation. The steam is generated here by way of example by combustion of natural gas.

    [0192] A thermodynamic simulation conducted by way of example gives the specific consumptions that follow, which take account both of the moistening of the gas substream with water and of the drying of the retentate or of the anode offgas downstream of the EHS with the aid of a propylene glycol scrub.

    [0193] For the generation of the steam, for every kg of water vapor volume introduced,


    K.sub.cal,H2O=2.29 kWh.sub.CH4/kg H2O  (6.4)

    of thermal energy is required. If this is provided by CH4 combustion, it is necessary for this purpose to burn, for every kg of water vapor volume introduced,


    K.sub.CH4,H2O=0.165 kg CH4/kg H2O  (6.5).

    [0194] This is used to calculate the hourly natural gas consumption for the moistening:


    m.sub.CH4,H2O=K.sub.CH4,H2O*m.sub.TG,H2O=14.0 kg CH4/h  (6.6)

    which is supplied to the steam generator via stream 22.

    [0195] For the operation of the glycol scrub,


    K.sub.el,H2O=0.97 kWh.sub.el/kg H2O  (6.7)

    of electrical energy is required for the refrigeration in the cold box. This is used to calculate the electrical power consumption for the moistening P3:


    P3=K.sub.el,H2O*m.sub.TG,H2O=82 kW.sub.el  (6.8)

    [0196] In total, for the moistening and drying of the gas substream, it is necessary to transfer


    K.sub.HT,H2O=7.99 kWh/kg H2O  (6.9)

    of heat. This is used to calculate the total heat transfer output Q.sub.HT,H2O


    Q.sub.HT,H2O=Q3+Q4+Q5+Q6=K.sub.HT,H2O*m.sub.TG,H2O=681 kW  (6.10)


    With


    K.sub.H2O=(M.sub.H2O/M.sub.H2)*[(Y.sub.TG,H2O,ω−Y.sub.TG,H2O,α)/Y.sub.PL,H2,α]*(1/AG)=2.13  (6.11)

    the specific consumption values based on the amount of product m.sub.H2,Pr are then found for natural gas demand:


    h.sub.H2O=m.sub.CH4,H2O/m.sub.H2,Pr=K.sub.cal,H2O*K.sub.H2O=4.87 kg CH4/kg H2  (6.12)


    for the power demand:


    p.sub.H2O=P3/m.sub.H2,Pr=K.sub.el,H2O*K.sub.H2O=2.06 kW.sub.el/kg H2  (6.13)

    and for the heat transfer output:


    q.sub.H2O=Q.sub.HT,H2O/m.sub.H2,Pr=K.sub.HT,H2O*K.sub.H2O=17.0 kW.sub.them/kg H2  (6.14)

    7) Separation Energy Expenditure

    [0197] The total separation energy expenditure TA.sub.tot is composed of a variable component TA.sub.var and a fixed component TA.sub.fix


    TA.sub.tot=TA.sub.var+TA.sub.fix  (7.1)

    [0198] The variable component based on the H2 product volume flow rate includes all the essential operating media


    ta.sub.var=TA.sub.var/m.sub.H2,Pr=h.sub.H2O*(1/F.sub.el,Gas)+p.sub.comp+p.sub.EHS+p.sub.H2O  (7.2)

    with the power-gas factor F.sub.el,Gas which describes the ratio of power price to natural gas price and for which, by way of example, the value


    F.sub.el,Gas=4.0 kW.sub.el/kW.sub.cal  (7.3)


    It follows from this that


    ta.sub.var=4.29/4.0+0.44+4.84+2.19=8,50 kW.sub.el/kg H2O  (7.4)

    [0199] The general formula for the ascertaining of the H2 product-specific variable separation energy expenditure is thus:


    ta.sub.var=(1/AG)*[K.sub.moist*[(Y.sub.TG,H2O,ω−Y.sub.TG,H2O,α)/Y.sub.PL,H2,α]+K.sub.comp*(1+1/Y.sub.PL,H2,α)*In[(p1+Δp)/p1)])+K.sub.F*{U.sub.EHS,over+K.sub.U,min*In[1+(1+Y.sub.TG,H2O,ω)/Y.sub.PL,H2,α/(1−AG)]}  (7.5)


    with


    K.sub.F=2*F/M.sub.H2/3600=26.59 Ah/kg H2  (7.6)


    and


    K.sub.moist=(M.sub.H2O/M.sub.H2)*(K.sub.cal,H2O/F.sub.el,Gas+K.sub.el,H2O)=13.78 kW.sub.el/kg H2  (7.7)


    and


    p1=pressure upstream of compressor  (7.8)

    [0200] The variation of AG, with equation (2.2), for each of the parameters gives an optimal gas subvolume m.sub.TG, α.sub.,op at which the H2 product-specific variable separation energy expenditure is at a minimum:


    m.sub.TG,var,α.sub.opt=(M.sub.PL/M.sub.H2)*(1/AG.sub.var,opt)*(1+1/Y.sub.PL,H2,α)*m.sub.H2,Pr  (7.9)

    [0201] FIGS. 3, 4 and 5 show, by way of example, the dependence of the H2 product-specific variable separation energy expenditure on the depletion level AG. The optimal separation level is always in the range between 0.80<AG.sub.var,opt<0.99.

    [0202] The fixed component of the necessary separation energy expenditure, based on the H2 product volume flow rate, derives from the capital costs. The capital costs are proportional to the power for the raising of the pressure p.sub.comp, the heat transfer output for the heating or cooling of the gas substream q.sub.TG, the power for the moistening and drying of the gas substream p.sub.H2O and q.sub.H2O, and for the separation performance of the EHS p.sub.EHS.

    [0203] Specific capital expenditure costs I.sub.spec are proportional to the sum total of all expenditures Σ(p.sub.i+q.sub.i).


    Σ(p.sub.i+q.sub.i)=p.sub.comp+q.sub.TG+p.sub.H2O+q.sub.H2O+p.sub.EHS

    [0204] FIGS. 6 to 8 show that there is also minimum for the sum Σ(p.sub.i+q.sub.i) that is of relevance in respect of capital costs. In other words, there also exists an optimal gas substream. The optimal separation level here is always in the range between 0.90<AG.sub.fix,opt<0.99.


    m.sub.TG,fix,α,opt=(M.sub.PL/M.sub.H2)*(1/AG.sub.fix,opt)*(1+1/Y.sub.PL,H2,α)*m.sub.H2,Pr

    [0205] The complex dependence of the optimal total depletion level AG.sub.opt—which takes account both of the variable and of the fixed component—on the two main influencing parameters of H.sub.2 loading of the natural gas stream in the pipeline Y.sub.PL,H2,α and steam loading of the gas substream before entry into the separation unit Y.sub.TG,H2O,ω, for the case that a pressure drop Δp of 0.1 bar occurs over the entire pathway of the gas portion from the withdrawal station 15 to the refeeding station 16, is listed in table 1 for some sampling points and is shown in FIG. 10.

    TABLE-US-00003 TABLE 1 AG.sub.opt as a function of Y.sub.H2, PL, α and Y.sub.TG, H2O, ω with Δp = 0.1 bar Y.sub.TG, H2O, ω Y.sub.H2, PL, α 0.005 0.010 0.020 0.025 0.030 0.040 0.050 0.100 0.150 0.200 0.010 0.931 0.960 0.975 0.983 0.985 0.988 0.992 0.996 0.997 0.999 0.020 0.879 0.926 0.960 0.967 0.972 0.979 0.984 0.991 0.994 0.995 0.025 0.855 0.912 0.950 0.960 0.967 0.974 0.979 0.990 0.993 0.995 0.030 0.836 0.899 0.942 0.953 0.960 0.969 0.975 0.987 0.991 0.993 0.040 0.804 0.874 0.926 0.939 0.948 0.959 0.967 0.983 0.988 0.991 0.050 0.775 0.852 0.911 0.926 0.936 0.950 0.959 0.979 0.985 0.989 0.100 0.675 0.769 0.849 0.871 0.887 0.910 0.925 0.959 0.972 0.978 0.150 0.618 0.714 0.803 0.829 0.848 0.876 0.896 0.941 0.959 0.968 0.200 0.579 0.673 0.766 0.795 0.817 0.848 0.870 0.925 0.947 0.959