On-Board and off-Board Performance of Hydrogen Storage Options

Monday, November 8, 2010: 4:36 PM
Alta Room (Marriott Downtown)
Rajesh K. Ahluwalia1, Thanh Q. Hua2 and J.-K. Peng2, (1)Nuclear Engineering Division, Argonne National Laboratory, Argonne, IL, (2)Nuclear Engineering, Argonne National Laboratory, Argonne, IL

A number of physical and materials-based hydrogen storage options are being developed to simultaneously meet the vehicular targets for gravimetric and volumetric capacity, cost, efficiency and greenhouse gas emissions, durability and operability, fuel purity, and environmental health and safety. The purpose of this paper is to evaluate the currently available physical, complex metal hydride, sorbent, and chemical hydrogen storage methods for their potential to meet these targets.

Our analyses show that hydrogen stored as a compressed gas (cH2) at 350 bar in Type III or Type IV tanks cannot approach the near-term (2010) volumetric target of 28 g/L. Increasing the storage pressure to 700 bar improves the volumetric capacity (without reaching the target value) but degrades the gravimetric capacity. Dormancy and hydrogen loss are major concerns if hydrogen is stored as a cryogenic liquid (LH2) in insulated low-pressure tanks, even if advanced vapor-cooled heat shield concepts are deployed. These problems can be mitigated by storing cryogenic hydrogen in pressure-bearing insulated tanks (the cryo-compressed option, CcH2). Our analyses confirm that CcH2 can also meet the near-term and intermediate-term (2015) targets for gravimetric and volumetric capacities. However, the cost of the cryo-compressed system is projected to be 3-4 times the target cost of $4/kWh, and the well-to-tank efficiency for the common pathway of producing hydrogen by central steam methane reforming (SMR) and subsequent hydrogen liquefaction is only about 41%, significantly lower than the efficiency target of 60%.

We have evaluated an on-board regenerable complex metal hydride (NaAlH4, sodium alanate) system and an off-board regenerable metal hydride (AlH3, alane) system. Our analyses show that, because of the large heat of reaction ( H = 37 kJ/mol-H2) and slow hydrogenation and dehydrogenation kinetics, the doped sodium alanate system needs a combustor for releasing hydrogen, a bulky heat exchanger for refueling in 3-5 min, and a hydrogen buffer tank to handle startup and rapid demand transients. On the other hand, our analyses show that alane is an attractive hydrogen carrier (10 wt% H2 content, H = 6.7 kJ/mol-H2), if it can be prepared and used as a slurry with >50% solids loading and a volume-exchange tank concept is developed that can store fresh and spent slurry in the same tank with a movable partition. Regenerating AlH3 is a major problem, however, since it is metastable and it cannot be directly regenerated by reacting the spent Al with H2. We have analyzed different indirect methods of regenerating AlH3, including the organometallic route. We have constructed a flowsheet for this three-step method in which AlH3 is formed as a stable adduct to an amine, the adduct is destabilized by transamination, and the AlH3 is released by the catalytic decomposition of the transaminated adduct by decomposing under vacuum. We concluded that fuel cycle efficiencies in 40-55% range are possible with this organometallic route, if low-grade waste heat is available for the decomposition steps and the electricity requirements for the vacuum distillation steps match the estimates.

We have analyzed two sorbent-based hydrogen storage systems, one using AX-21, a high surface-area superactivated carbon, and the other using MOF-177, a metal organic framework material, Zn4O(1,3,5-benzenetribenzoate) crystals. We found that moderate pressures and cryogenic temperatures are needed for reasonably high storage capacities, and that a temperature swing must be imposed in addition to the pressure swing to desorb hydrogen. Using off-board liquid N2 for removing the heat of sorption is not very attractive because of the poor thermal conductivity of the sorbent bed and the unfavorable fuel cycle (well-to-tank) efficiency. We also evaluated the adiabatic refueling option, in which the sorbent bed is cooled evaporatively by charging it with liquid hydrogen, and warmed hydrogen is recirculated during discharge to supply the heat of desorption. The results from our analyses indicate higher storage capacities with MOF-177 than with AX-21, but these capacities are lower than those available with the cryo-compressed option.

We have also evaluated several exothermic and endothermic chemical methods of storing hydrogen. Releasing hydrogen by hydrolysis of sodium borohydride (NaBH4, SBH) has been demonstrated in the laboratory and on-board an experimental vehicle. Our analysis of the SBH system pointed to a number of shortcomings, including the limited solubility of the product of reaction (NaBO2) in the aqueous alkaline medium, difficulty in rejecting the large amount of heat that is liberated in the exothermic reaction ( H = 52.5 kJ/mol-H2), and the low well-to-tank efficiency in regenerating NaBH4 from NaBO2. We concluded that, for reasonable fuel cycle efficiencies, formation of BO bonds should be avoided. We have evaluated the option of using an organic liquid carriers (LCH2), such as N-ethylcarbazole, which can be dehydrogenated thermolytically on-board a vehicle and rehydrogenated efficiently in a central plant by well-established methods and processes. Our analysis indicated low gravimetric (<3.5 wt% H2) and volumetric (<25 g/L) capacities of this system, due to large H (51 kJ/mol-H2), pressure-limited equilibrium conversion, and slow dehydrogenation kinetics. As a result, the on-board system requires a mineral oil-based high-temperature coolant, a combustor, and a buffer H2 storage tank for startup and rapid transients. Using reverse engineering, we have also determined the attributes that a prospective organic liquid carrier must have for such an on-board system to meet the H2 storage targets.

Recently, there has been much interest in ammonia borane (NH3BH3, AB) as a hydrogen carrier because it has a high H2 content (19.6 wt%), it undergoes exothermic thermolysis to release the H2 ( H = 33 kJ/mol-H2), and it decomposes at relatively low temperatures. Like AlH3, AB is thermodynamically unstable and its decomposition rate is independent of the backpressure. We are evaluating the concept of using AB dissolved in an organic liquid solvent (1-butyl-3-methylimadazolium, bmimCl) as an on-board storage medium. We have devised reactor configurations for controlling the peak temperature through heat transfer and recycle of the product stream. Like AlH3, regenerating AB efficiently from the spent products is particularly challenging; it would be quite impractical if all three equivalents of H2 are released from AB, allowing boron nitride (a ceramic) is allowed to form. We have constructed and analyzed flowsheets for off-board regeneration of spent AB using process chemistry developed by others. We have identified process steps that are energy intensive and are the largest contributors to inefficiencies if thiol or an alcohol is used to digest the spent AB. We concluded that, on a primary energy basis, regeneration efficiencies ranging from 17 to 35% are possible with these digestion schemes.

Figure 1 summarizes the volumetric and gravimetric capacities of the different on-board hydrogen storage options evaluated to date and compares them with the near-term, intermediate-term and ultimate targets. Figure 1 also compares the well-to-engine (WTE) efficiencies of these options for the pathway in which H2 is produced in a central SMR plant. The paper will discuss others aspects of the storage options as well, including high-volume manufacturing cost projections, greenhouse gas emissions, durability and operability, and environmental health and safety issues (dormancy and hydrogen loss).

Figure 1 System capacities and fuel cycle efficiencies of different hydrogen storage options

Extended Abstract: File Not Uploaded