Design and Control of Stand-Alone Autothermal Reforming Systems Using Methanol

Abstract The conceptual design of compact , stand-alone autothermal reforming systems using methanol is presented. Based on specific conditions of maximum waste heat recovery , the optimization algorithm for maximizing hydrogen yield is obtained. Through the closed-loop control tests for the Aspen Dynamics model , the feasible manipulation by adjusting preheated water flow is established and the stable output regulation of nonlinear systems in the presence of unknown inlet perturbations is guaranteed. Introduction      Methanol has certain advantages compared to other hydrocarbons due to the low reforming temperature , ease of handling and a relatively high hydrogen-carbon ratio. Similarly , ethanol is also a potentially attractive feedstock because of its availability , nontoxicity and handling safety. From the aspect of kinetics , the reforming processes using different catalysts could affect the hydrogen yield , conversion efficiency and selectivity of undesired products. Hung et al.1 provided an optimal operating conditions by adjusting H2O/ethanol and O2/ethanol ratios to improve catalytic activity. In terms of catalyst utilization , Tang et al.2 showed that the autothermal reformer has superior performance over steam reformer. The stand-alone fuel cell power system associated with fuel processing units for stationary and potentially mobile applications was recently investigated. Wang and Wang9 showed that the autothermal reforming system is optimized through thermodynamic equilibrium and exergetic analyses. Stamps and Gatzke3 showed that a PEMFC stack coupled to methanol reformer with aid of a specific auxiliary power could effectively meet desired power targets. Degliuomini et al.4showed the dynamic manipulation and control of a heat integrated PEMFC power system connected to ethanol fueled processing units.      In this work , stand-alone autothermal reforming systems using methanol is evaluated according to hydrogen yield , cost of hydrogen production and energy consumption. As for reducing fuel consumption as well as increasing heat recovery capability , the optimization algorithm for maximizing hydrogen yield of autothermal MeOH-to-H2 processor subject to the condition of maximum waste heat recovery is performed. Moreover , the dynamic manipulation and control implementation is deployed in Aspen Dynamics environment. According to the closed-loop autotune variation (ATV) tests of the whole dynamics system in the face of unknown inlet perturbations , the PID control method can ensure the no offset and stable output regulation performance. Autothermal MeOH-to-H2processor Autothermal alcohols-to-H2processor         In Fig. 1 , a so-called autothermal alcohols-to-H2 processor is considered as another stand-alone hydrogen production process , where an autothermal reformer (ATR) is an adiabatic plug flow reactor. It is noted that the flue gas from the burner is directly used to adjust the inlet temperature of the ATR , . Regarding the heat recovery design for autothermal alcohols-to-hydrogen processors in Fig. 1 , the water flow is preheated by the outlet stream of the WGS reactor using a heat exchanger (E-102) , and the feed flow is preheated by the high-temperature flue gas from the burner using anther heat exchanger (E-101). Notably , the heating jacket is removed and the inlet temperature of the ATR , TATR ,in , can be adjusted.                                                        Fig. 1 Autothermal alcohols-to-H2processor Optimization      To address the maximum waste heat recovery , the following optimization algorithm for maximizing TATR ,in .  Assumed that TATR ,in is adjusted by changing the heat capacity of exchanger (UA) , and UAurepresents an ultimate bound of UA. If the maximum inlet temperature of the ATR is achieved , i.e. , then the corresponding  would be close to the minimum value. Furthermore , the optimization algorithm for maximizing the hydrogen  yield. Fig. 2 shows that the maximum hydrogen yield is obtained while the feed ratio (S/C) and  are specified at 0.75 and 833.9K , respectively. Figure 2 Sensitivity analysis      Based on above optimization algorithms , the steady-state simulation under optimal operating conditions is shown in Fig. 3. Notably , it can ensure a lowest flue gas temperature , e.g. Tflue ,out=321K. Figure 3   Optimization and steady-state simulation Process control      Aspen Dynamics are employed to model the system dynamics with prescribed steady-state initials. To address a feasible dynamic manipulation , if the initial temperature of TATR ,inis 799.4 K and the corresponding UA of heat exchanger (E-101) is fixed at 463.2 K/W , then the open-loop tests of the system with respect to  step changes of water flow are shown in Fig. 4. Figure 4 Open-loop tests of system with  step changes of water flowrate      It shows that the manipulation of water flow not only dominates the hydrogen yield but also it affects the waste heat recovery capability. Moreover , a closed-loop ATV test is used to determine the ultimate period and gain when the hydrogen production rate and water flow rate are treated as the controlled output and manipulated input , respectively. The default value of the relay output amplitude is 5% , which is usually good. Fig. 5 shows that ATV test is finished after several (4–6) cycles. Fig. 6 shows that the PID controller settings are determined by Ziegler–Nichols tuning rule. Figure 5 Closed-loop ATV test  Figure 6 PID controller settings To check the disturbance rejection capabilities , Fig. 7 shows that the stable and no offset output regulation is achieved while  step changes of inlet temperature of water flow appear. Figure 7  perturbations in the inlet temperature of water flow By steady state situation , the higher inlet temperature of water flow can induce the higher TATR ,in but Tflue ,outis not affected. Conclusions      With respect to hydrogen yield , cost of hydrogen production and energy consumption , the simulation shows that the autothermal MeOH-to-H2 processor can effectively maximize hydrogen yield as well as minimize heat loss. To address feasible dynamic manipulation , the PID control design using closed-loop ATV tests is employed. Although the autothermal MeOH-to-H2processor is a typically nonlinear system due to complex kinetics and specific heat recovery conditions , the single-input and single-output (SISO) control structure is successfully established by Aspen Dynamics. Simulation shows that the disturbance rejection by manipulating a preheated water flow can be guaranteed. References (1)     Hung , C. C.; Chen , S. L.; Liao , Y. K.; Chen , C. H.; Wang , J. H. Oxidative steam reforming of ethanol for hydrogen production on M/Al2O3. Int. J. Hydrogen Energy , 2012 , 37 , 4955. (2)     Tang , H.-Y.; Erickson , P.; Yoon , H. C.; Liao , C.-H. Comparison of steam and autothermal reforming of methanol using a packed-bed low-cost copper catalyst. Int. J. Hydrogen Energy , 2009 , 34 , 7656. (3)     Stamps , A. T.; Gatzke , E. P. Dynamic modeling of a methanol reformer—PEMFC stack system for analysis and design. J. Power Source , 2006 , 161 , 356. (4)     Degliuomini , L. N.; Biset , S.; Luppi , P.; Basualdo , M. S. A rigorous computational model for hydrogen production from bio-ethanol to feed a fuel cell stack. Int. J. Hydrogen Energy , 2012 , 37 , 3108. (5)     Amphlett , J. C.; Creber , K. A. M.; Davis , J. M.; Mann , R. F.; Peppley , B. A.; Stokes , D. M. Hydrogen production by steam reforming of methanol for polymer electrolyte fuel cells. Int. J. Hydrogen Energy 1994 , 19 , 131. (6)     Choi , Y. , Stenger , H. G. Water gas shift reaction kinetics and reactor modeling for fuel cell grade hydrogen. J. Power Sources 2003 , 124 , 432.