GENHSTOK: more pure hydrogen production with CO 2 capture

Transcription

1 WHEC 16 / June 006 Lyon France GENHSTOK: more pure hydrogen production with capture P. arty a, D. Grouset a, S. Lecoq a, J.-C. Hoguet a a N-GHY S.A. - ZI ontplaisir - Site Industriel Saint Antoine 51, rue Isaac Newton Albi - FRANCE n-ghy@n-ghy.com ABSTRACT: In 005, the Hot Section of a 100 kw th high pressure hydrogen generator from Diesel Fuel, pure oxygen and water has been designed, realised, tested and CE certified. This demonstration is the initial key step of the GENHSTOK process for high pressure pure hydrogen production from any fossil fuel or bio-fuel without any emission. The cumulated duration of the early tests is 50 hours. Even in operating conditions far from nominal conditions, this hydrogen generator shows good performances (nearly 73 % H efficiency). This paper presents some tests results in steady-state conditions. ass and energy balances are given in details. KEYWORDS: hydrogen production, capture Background Fuel cell based technologies currently know tremendous developments for different future applications such as electric vehicles or domestic/industrial cogeneration. ost fuel cells run on pure hydrogen but free hydrogen does not exist in the nature and has to be produced: - Either in centralised or decentralised production units for further delivery to retailers and local users - Or locally, just previous to the fuel cell, for immediate consumption. Two different hydrogen sources can be distinguished: - Hydrogen produced from downstream by water decomposition: thermally at high temperatures or electrically by electrolysis. Given that hydrogen will generate electricity in a fuel cell, the electrolysis pathway may seem to be an aberration, at least on the global energetic efficiency point of view. If the required electricity is produced from renewable sources (wind, solar, geothermal) or from nuclear source, at the very most there is no or other pollutant formation along with this production/consumption chain. Hydrogen clearly appears in this case as an energy carrier enabling clean electricity production through fuel cells in places (or at times) that are short of nuclear or renewable energy, for stationary or mobile applications indifferently. - Hydrogen produced from upstream by hydrocarbon reforming. Fossil hydrocarbon reforming generates hydrogen but also, which weakens the interest in using fuel cells. However this pathway advantageously offers potentially high energy efficiencies that could contribute to savings of fossil fuels or that could enable the use of biomass productions for energy-oriented purposes. Thus there is an emerging need for big centralised or small local hydrogen production units with good energy efficiencies and with little side productions of and other pollutants. In addition in many industrial sites, pure or nearly pure oxygen is produced without further valorisation: - In hydrogen production plants using the water electrolysis process, where an oxygen flow is generated at the cathodes - In air distillation plants producing nitrogen for passivating or refrigerating purposes, where nitrogen and pure or nearly pure oxygen are produced in the liquid form. N-GHY S.A., Fuel Processors for Fuel Cells, is developing devices able to produce pure hydrogen from heavy fuels with pure oxygen and water as oxidants. The N-GHY team has been developing for more than eight years a unique process of hydrogen generation from heavy fuels. This breakthrough process requires no reformer catalyst to achieve high performances. The so-called High Temperature Hybrid Steam- Reforming (HT HSR) technology is simple, fuel-flexible, and particularly well tailored to the above mentioned applications. 1/1

2 WHEC 16 / June 006 Lyon France Decentralised pure H production with capture: N-GHY s strategy Analysis of H production and transport Here are the main conclusions of N-GHY which agree with those of most experts on H production and transport: - transports of liquid and gaseous H by truck are expensive options - onsite water electrolysis for H production is a suitable pathway only for small production units and will necessitate the availability of green, excess electricity - the cheapest way of producing H is hydrocarbons reforming - H pipelines will not reach immediately all H consuming sites and it will last decades before the deployment of the required network is completed - A market sector for H currently exists: the market of industrial H, in the 30 to 1000 Nm 3 /h range ( 100 kw to 3 W) e.g. for: constitution of reducing atmospheres, electronic applications, glassmaking - A market also exists in different sectors: food (gaseous drinks, refrigeration/deepfreezing/preserving of food), agriculture (greenhouses), waste water treatment, industry (as a propulsive agent replacing CFCs, as an expanding agent for foams, as an extinguishing agent, as a solvent in its supercritical state ) N-GHY is convinced that: - it is necessary to have a technology of decentralised hydrogen production, from hydrocarbons - capture and bio-fuel use are important advantages GENHSTOK patented process The work presented in this paper is related to the GENHSTOK patented process for hydrogen storage and delivery under pressure, especially in refuelling stations. Pure H production involves a H / separation step, via a PSA or metallic membrane. The path studied by N-GHY makes use of a high pressure Diesel fuel processing unit fed with pure O. Fuel Conversion to H + to conversion Permeation membrane HP H storage Oxygen Liquid tank GENHSTOK process for pure H production and storage Key assets of the GENHSTOK patented process are as follows: - The use of pure O leads to a significantly higher efficiency compared to the use of air: conversions are complete and energy efficiencies higher than 85% can be reached - As a non-catalytic process, GENHSTOK is suitable for any type of fuel: natural gas, LPG, fuel oil, naphtha or heavier fuels and even bio-fuels (vegetable oil, methyl ester vegetable oil, bio-ethanol, biogas ) - Permeation membrane is very efficient, due to a high H concentration (no dilution by N ) and a high driving force (resulting from the high pressure of the reformate) - GENHSTOK offers the possibility to capture and store : high pressure operation (typically 60 bars) leads to an easy condensation and a liquid compact storage for further valorisation or sequestration. Thus, if H comes from a fossil hydrocarbon, then its production does not generate any additional greenhouse gases and if it comes from a bio-fuel, then its production ends in a reserve. - GENHSTOK can be advantageously coupled with water electrolysis. The production of 1 kg of electrolytic H goes together with the production of 8 kg of O. Electrolysis can be operated at high /1

3 WHEC 16 / June 006 Lyon France pressure or else, as the H, this O has to be compressed for further valorisation. Such a valorisation for fuel reforming should be preferentially done on site, because the chemical H flow will be added to the electrolytic one while contributing to sequestration and to the paying-off of the different compressors and utilities used for the conditioning of the produced H. The lever effect is important since with the 8 kg of O that are produced by water electrolysis and considering a reasonable conversion efficiency of 80% to 90% for the H production step, the Diesel fuel consumption will vary from 5.75 to 6.90 kg / kg of electrolytic H and the quantity of sequestered from 18.1 to 1.8 kg / kg of electrolytic H. The amount of generated H will be in the range of 1.5 to 1.96 kg of chemical H / kg of electrolytic H. By coupling both processes together, the increase in H production is finally of the order of.5, which greatly compensates the efficiency losses of the electrolytic process. Bio-fuel or fossil fuel 5.7 kg (Natural gas) 6.3 kg (Diesel fuel) Electricity 9 kg GENHSTOK process O storage 8 kg High Pressure Electrolyser Liquid tank 15.7 kg (from Natural gas) 0 kg (from Diesel fuel) HP H storage 1.5 kg (from Natural gas) 1.7 kg (from Diesel fuel) 1 kg (from electricity) GENHSTOK process coupled with water electrolysis for pure H production and capture 100 kw th / 55 bars diesel fuel reformer demonstration Introduction As STEP0 towards the real scale 1W H production unit, reforming tests were successfully conducted in 00 without preheating of the reactants on a test bench rating a thermal power of 5 kw at pressures up to 6 bars, with atmospheric air and with O enriched air. Liquid and gaseous fuels testings were accomplished. Fuels tested were: propane, methane, ethanol, rapeseed oil methyl ester, crude rapeseed oil, sunflower oil methyl ester and commercial Diesel fuel (available in current refuelling stations). Reforming tests results were well consistent with theoretical computations. Obtained H + conversions and energy efficiencies were near the equilibrium values computed in adiabatic conditions. The nearly adiabatic conditions and the rise in the reaction temperature both resulted in reformates with extremely low contents of unreacted hydrocarbons, within short residence times (below 0.1 s). So, as STEP1 towards a real, pre-industrial scale H production unit, a demonstration reforming reactor design was completed early 005. It has been assembled and tested in N-GHY facilities. It has been CE certified in 005. This demonstration unit consumes commercial Diesel fuel (directly available in today s refuelling stations), liquid water and pure O stored in compressed tanks. It produces a clean H + rich gas mixture. Its maximum operating power is greater than 100 kw th (based on the HHV of H ) under 55 bars. This power is representative of the power required for a H refuelling station feeding a 3 buses depot (100 kw th H = 8. Nm 3 /h H = 60.8 kg/day H ). This prototype is equipped with a fully numerical safety, control and monitoring system. Its design includes a high grade thermal integration: high temperature preheating and low external losses lead to a high efficiency (85% nominal). Numerous sensors are used for safety, performance evaluation and equipment qualification of this pure O /commercial Diesel fuel High Pressure Hybrid Steam- Reformer demonstrator. The demonstration unit and some tests results are presented below. 3/1

4 WHEC 16 / June 006 Lyon France 100 kw th / 55 bars Diesel fuel reformer Hybrid Steam-Reforming principle Hybrid Steam-Reforming (HSR) is a combination of two reactions: 1. Combustion of a part of the fuel in oxygen (exothermic). Steam-reforming of the remaining part of the fuel (endothermic) The heat generated by the exothermic combustion is used to realise the endothermic steam-reforming reaction ending in the production of a hydrogen rich mixture. EO H O ENDO O BUSTION f of O O mole of fuel STEA-REFORING of (1- f O ) mole of fuel H -rich mixture 1 mole of fuel HSR principle Depending on the global oxygen factor (f O ), the combination of both reactions can be exo-, endo- or autothermic, which explains why the global reaction is called hybrid steam-reforming. For a given fuel F, the theoretical H efficiency (HHV based) that can be reached is only a function of f O : 1 f max O η H F =, 1 ν F ν F depends on the chemical composition of the fuel and on its enthalpy of formation. For example, its value is.1% for Diesel fuel. When f O = ν F, the theoretical efficiency is 100%, corresponding to ideal auto-thermal reforming (ATR). The specificity of N-GHY s process, compared with other competing technologies, is that it runs at high temperature, typically 100 C. ain advantages of this specificity are as follows: - the use of reforming catalysts is not necessary - full conversion of the fuel is reached - the process is adaptable to any king of liquid or gaseous fuel: natural gas, LPG, fuel oil, naphtha, diesel fuel, kerosene, bio-fuels /1

5 WHEC 16 / June 006 Lyon France Description of the Hot Section The so-called Hot Section consists in three parts: - the reformer, constituted by : o the reactor itself (or reacting chamber) where both combustion and reforming reactions occur (ending in the production of H,, ) o and by internal reactants heaters and super-heaters - two external heat exchangers to preheat water: HE1 and HE Schematic description of the Hot Section Reformate The reformate (H / / mixture) is generated in the reactor (reacting chamber). It is collected at the exit of the reactor in a circular manifold and then goes through the two external heat exchangers downstream (HE, HE1) where process water is preheated. Oxygen Gaseous oxygen is stored in pressurized bottles (00 bars). It is directly injected into the combustion chamber without any preheating. Diesel fuel A pump is used to pressurize liquid Diesel fuel. For start-up, diesel fuel is first injected directly in the combustion zone within a commercial pressure atomiser, without any preheating. When the reformer is operated in nominal conditions, Diesel fuel is only introduced in the reactor through a central thermally insulated tube situated at the bottom of the reformer. At the end of this tube, Diesel fuel is mixed with and vaporized by superheated steam. The temperature of the mixture is 50 C, enabling complete vaporization of the Diesel fuel. Then the mixture is collected at the bottom of the reformer by a manifold and directed towards a super-heater that raises the temperature of the mixture to 700 C. Finally this hot mixture of Diesel fuel and water vapours reacts with oxygen in the reaction chamber where reformate production begins immediately. As for Diesel fuel, a pump is used to pressurize liquid water. For start-up, cold liquid water is sprayed in the combustion zone. For nominal operation, water is not introduced in the combustion zone anymore: the only water introduced is super-heated. The circuit used for heating the water is as described below: 5/1

6 WHEC 16 / June 006 Lyon France - a 1 st external heat exchanger (HE1) preheats liquid water - then preheated liquid water flows through a cylindrical double wall within the reformer, where the water is vaporized (temperature around 70 C at 55 bars) - a nd external heat exchanger (HE) is then used to superheat the steam - after HE, steam is collected in a circular manifold and is sent to a final internal super-heater which raises the steam temperature to 700 C. temperatures at the outlet of the double wall and at the outlet of the final steam super-heater are automatically adjusted by two temperature regulation loops (each constituted by a thermocouple and a 3-way valve enabling the partial or total bypass of the external HE situated downstream). External heat exchangers Liquid cooling loop In order to cool the hub cap of the combustion chamber that is made of steel, a water cooling loop is used. Test results The Hot Section was tested in N-GHY s facilities from September 005 to January 006. The cumulated duration of the first series of tests is 50 hours. The results that are discussed below are related to a nearly 1-hour test. For simplicity and clarity reasons, comments are focusing on a 1-hour time segment, during which the reformer is operated in steady-state conditions. Operating conditions The table below gives the operating conditions of the test. Operating conditions DESIGN TEST (nominal) Power (HHV Diesel fuel) kw th 33.6 kw th Oxygen factor (f O ) factor (f w ) Pressure 55 bars abs bars abs. Direct liquid Diesel fuel injection (in the combustion zone) NO YES Direct liquid water injection (in the combustion zone) NO YES Test pressure and experimental water and oxygen factors are very close to design values. The system is designed to reach an hydrogen efficiency of 85.0% (HHV based) at the nominal oxygen factor (0.355). The mean experimental oxygen factor (0.379) is slightly above the nominal one. Thus, the maximum theoretical H efficiency that can be reached in this test is 81.8%. Test conditions differ from design values for the following reasons: - first of all the experimental operating power is 33.6 kw th, which is approximately a third of the design power (115.8 kw th ) - some liquid water and some liquid Diesel fuel are directly introduced in the combustion zone, i.e. without preheating: this lowers the maximum temperature and performance that can be reached All reactants mass flows are given in the table below. 6/1

7 WHEC 16 / June 006 Lyon France Reactant ass flow (g/s) Diesel fuel (combustion) 0.30 Oxygen 0.90 (combustion) Diesel fuel (reforming) (reforming) 1.10 Figures below show the constancy of the different operating parameters during the test. 0,5 Power : 33.6 kw th (HHV Diesel fuel) 30 factor (f w ) : 1.56 Power (kw th ) - f w 0 1, Oxygen factor (f O ) : ,5 f O 0 0 :30 60 :0 :50 3:00 3:10 3:0 3:30 Time Steam 55 Pressure (bars - abs.) 50 5 Reformate (reforming zone) Reformate (upstream from PV) Order given to the reformate PV 0 :30 :0 :50 3:00 3:10 3:0 3:30 Time Operating parameters Gas analysis A mass spectrometer (AETEK Dycor Proline) was used to realize the gas analysis. The evolutions of the measured volume fractions (H,,, ) are drawn below. ean values are also given above the curves. The measured composition is nearly constant over the time period. 7/1

8 WHEC 16 / June 006 Lyon France H : 51.1% Dry volume fractions (%) : 9.3% : 37.3% :.% 0 :30 :0 :50 3:00 3:10 3:0 3:30 Time easured volume fractions Atomic balances Atomic balances are drawn up on the following assumptions: - the chemical formula of Diesel Fuel () is C n H m O p with n = 13., m =.876, p = the volume fractions that are measured by the mass spectrometer are given on a basis -, and are the only C-containing gaseous compounds in the reformate, the formation of C or of heavier hydrocarbons than being taken into account through the introduction of the k C factor (see below) - gas leaks are negligible - the O balance is arbitrarily set : k = O Those assumptions are used to write: - the total mass flow conservancy: comb. ref comb. ref. m & = + + H O + H O + - the carbon mass flow conservancy: comb. ref. m & + + = kcn with k and to calculate: - the water mass flow contained in the reformate: comb. ref comb. ref = H O + C ( ) k C n H O H H H O + O + O comb + ref + - H,,, mass flows: H H = H = H O O = = H O H O - And finally C, H, O atomic balances (x atomic balance is defined as the ratio between the mass flow of atom x in the and the mass flow of atom x in the reactants): 8/1

9 WHEC 16 / June 006 Lyon France H H O H H O C = H = O = k comb ref comb ref comb ref + H O + H O + n + m Atomic balances are drawn on the figure below. ean values are close to 1.00, revealing acceptable gas analysis: C = 0.97, H = 1.07, O = 1.00 (assumption). H O O 1,0 1,10 1,00 0,90 0,80 0,70 0,60 0,50 0,0 0,30 0,0 H : 1.07 O : ,10 :30 :0 :50 3:00 3:10 3:0 3:30 Time Heure Atomic balances C : 0.97 ass balance To establish the energy balance of the hot section, the mass balance has to be set up first. Inlet mass flows (reactants) are known within a good accuracy. oreover, the method that is used by the mass spectrometer shows that we can trust the H and volume fractions. easured and calculated compositions are compared on the figure below. 3 successive results of calculations are proposed. The 1 st calculation is done on an adiabatic equilibrium hypothesis, with a preheating at 700 C. In this case, the adiabatic temperature is 130 C (which is lower than the measured temperature: 1350 C). Calculated H volume fraction (53.17%) is close to the measured one (51.07%). The sum of calculated and volume fractions (31.85% % = 6.70%) is also very close to the sum of measured and volume fractions (37.9% + 9.8% = 6.57%), but experimental / ratio is shifted towards persistence in comparison with theoretical / equilibrium. oreover, the measured volume fraction (.36%) is higher than the calculated one (0.03%). Given this very low theoretical content, the calculated H efficiency in adiabatic conditions equals the theoretical value calculated for f O = (81.8%). The nd calculation is done in order to take into account the effective persistence of. The composition obtained with the 1 st calculation is shifted towards persistence through reverse methane steamreforming reaction ( + H + H O). Given that this reverse reaction is exothermic, the computed temperature is higher (1306 C). The H content decreases (9.78%) and the associated H efficiency is lower (7.9%). The 3 rd calculation finally shifts the WGS equilibrium ( + H O + H ) towards H formation. Actually nd and 3 rd calculations are done iteratively until calculated H and volume fractions fit in with chemical measures (51.07% and.36%, respectively). As the WGS reaction is also exothermic, the final temperature (1316 C) is still slightly higher than previously. This temperature is also close to the temperature effectively measured in the reforming zone. Final and volume contents are consistent with WGS equilibrium at 1100 C. H efficiency is not affected by the 3 rd calculation. The assumed mass balance for energy balance set up is based on the results of this 3 rd calculation. Corresponding calculated mass flows are given on the figure below. 9/1

10 WHEC 16 / June 006 Lyon France Inlet Outlet Calculations easures Adiab. () Equil. shift Equil. Persistence WGS g/s Hot Section H g/s H O g/s 53.17% 9.78% 51.07% 51.07% O 0.90 g/s H O g/s ass Balance g/s 31.85% 31.76% g/s 1.85% 16.0% 8.56% 37.9% 18.07% 9.8% 0.01 g/s 0.03% C 1.00 H 1.00 O C η=81.8%.% C 1.00 H 1.00 O C η=7.9%.36%.36% C 1.00 H 1.00 O C η=7.9% C 0.97 H 1.07 O C ass balance (Hot Section) The persistence of in the reformate is mainly due to: - the too low temperature in the reaction chamber (1316 C instead of a design temperature of 100 C): at this temperature, the residence time is not long enough to reach the equilibrium - the injection of cold (not preheated) water and fuel directly in the reaction chamber - the heat losses Energy balance To set up the energy balance, a number of measures are taken into account. These are mentioned on the figure below. Hot Section Thermal losses (cooling water).6 kw Thermal losses (walls) 1.0 kw Diesel Fuel B H O B O B Diesel Fuel Reformer Reactor Super-heater 700 C Steam super-heater Reformer outlet 60 C H 51.07%/.36%/ HE HE1 HE1 outlet 80 C P = kw f O = f e = Double wall p = 50.5 bars H O easures taken into account for energy balance set up Thermal losses were evaluated from: - cooling water temperatures and flow measurements - walls temperatures measurements Assuming H, and mass flows given on the above figure, the overall energy balance of the Hot Section can be drawn up (see figure below). The obtained total power at his outlet (chemical power + sensitive power + thermal losses) is 3.86 kw. Comparing this value with the inlet power contained in the Diesel Fuel flow (33.61 kw), the calculated balance deviation is only 3.6%, which is rather low and acceptable. 10/1

11 WHEC 16 / June 006 Lyon France Inlet 5 C Diesel Fuel Power = kw HHV Hot Section η max, theoretical 81.8% η H HHV 7.9% Conversion 89.1% H (HHV) (HV) (HHV) Outlet 80 C kw 8.71 kw.7 kw Sensitive power.7 kw Balance deviation 3.6% Overall energy balance (Hot Section) Power loss (cooling) Power loss (walls) TOTAL.6 kw 1.0 kw 3.86 kw Figures below show the results of energy balances of HE1 and. All mentioned temperature values were effectively measured over the focused time period. Considered steam/water and reformate mass flows are related to real measured flows and the assumption is still made that there are no gas leaks. Furthermore, heat losses in heat exchangers are neglected. Concerning HE, on the reformate side, the calculated exchanged power is.8 kw. To make the exchanged power on the steam side fit in with this value (energy balance), the mass fraction of liquid water at the inlet of saturated vapour has to be adjusted to 85%. Regarding HE1, the calculated exchanged power on the reformate side (0.8 kw) is very close to the calculated exchanged power on the liquid water side (0.30 kw). Thus energy balance of HE1 is very good. 71 C 85% liquid Inlet Steam 1.1 g/s Reformate Inlet HE 3.19 g/s 3.19 g/s Exch. power (reformate):.8 kw 60 C Exch. power (steam):.7 kw 35 C Reformate Outlet 91 C 1.1 g/s Outlet Steam Energy balance (HE) 38 C Inlet 1.1 g/s Reformate Inlet HE g/s 3.19 g/s Exch. power (reformate): 0.8 kw 35 C Exch. power (water): 0.30 kw 80 C Reformate Outlet 10 C 1.1 g/s Outlet Energy balance (HE1) 11/1

12 WHEC 16 / June 006 Lyon France Last figure offers a synoptic view of the energy balance of the Hot Section. Power and temperature levels are mentioned. Diesel Fuel Power = kw HHV Hot Section (1110 C) 60 C 5 C 130 C 1306 C 1316 C 80 C 700 C 5.18 kw 6.99 kw HE1 : 0.30 kw HE :.7 kw Double wall + Super-heaters HE1 Double wall HE Steam super-heater Steam/Diesel Fuel vapour super-heater 1.65 kw 0.61 kw Cooling water.6 kw 1.0 kw Wall losses 1.0 kw Energy balance: global results (Hot Section) Conclusions and future work In 005, the Hot Section of a 100 kw high pressure hydrogen generator from Diesel fuel, pure oxygen and water has been designed, realised, tested and CE certified. The cumulated duration of the first series of tests is 50 hours. Even in operating conditions far from nominal conditions foreseen for its design, this hydrogen generator shows good performances (nearly 73 % H efficiency). Other validation tests are currently done. Efforts are mainly focused on improvements of injections and mixtures, thermal losses reductions and enhancement of heat exchanges. The temperature will raise in the reaction chamber to the design value so that equilibrium can be achieved. Nominal performance (85% H efficiency / 55 bars) should be reached before the end of 006. This demonstration is the initial key step of the GENHSTOK process for high pressure pure hydrogen production from any fossil fuel or bio-fuel without any emission. Next step is a 30 Nm 3 /h fully integrated system to be commercialised in rd step for N-GHY is a 500 Nm 3 /h unit to be commercialised in /1