In this thesis, ammonia (NH3) decomposition was assessed as a fuel processing
technology for producing on-demand hydrogen (H2) for portable and distributed fuel cell
applications. This study was motivated by the present lack of infrastructure to generate H2 for
proton exchange membrane (PEM) fuel cells. An overview of past and recent worldwide
research activities in the development of reactor technologies for portable and distributed
hydrogen generation via NH3 decomposition was presented in Chapter 2. The objective was to
uncover the principal challenges relating to the state-of-the-art in reactor technology and obtain
a basis for future improvements. Several important aspects such as reactor design, operability,
power generation capacity and efficiency (conversion and energy) were appraised for innovative
reactor technologies vis-à-vis microreactors, monolithic reactors, membrane reactors, and
electrochemical reactors (electrolyzers). It was observed that substantial research effort is
required to progress the innovative reactors to commercialization on a wide basis. The use of
integrated experimental-mathematical modelling approach (useful in attaining accurately
optimized designs) was notably non-existent for all reactors throughout the surveyed openliterature.
Microchannel reactors were however identified as a transformative reactor technology
for producing on-demand H2 for PEM cell applications.
Against this background, miniaturized H2 production in a stand-alone ammonia-fuelled
microchannel reactor (reformer) washcoated with a commercial Ni-Pt/Al2O3 catalyst (ActiSorb®
O6) was demonstrated successfully in Chapter 3. The reformer performance was evaluated by
investigating the effect of reaction temperature (450–700 °C) and gas-hourly-space-velocity
(6 520–32 600 Nml gcat
-1 h-1) on key performance parameters including NH3 conversion, residual
NH3 concentration, H2 production rate, and pressure drop. Particular attention was devoted to
defining operating conditions that minimised residual NH3 in reformate gas, while producing H2
at a satisfactory rate. The reformer operated in a daily start-up and shut-down (DSS)-like mode for a total 750 h comprising of 125 cycles, all to mimic frequent intermittent operation envisaged
for fuel cell systems. The reformer exhibited remarkable operation demonstrating 98.7% NH3
conversion at 32 600 Nml gcat
-1 h-1 and 700 °C to generate an estimated fuel cell power output of
5.7 We and power density of 16 kWe L-1 (based on effective reactor volume). At the same time,
reformer operation yielded low pressure drop (<10 Pa mm-1) for all conditions considered.
Overall, the microchannel reformer performed sufficiently exceptional to warrant serious
consideration in supplying H2 to low-power fuel cell systems.
In Chapter 4, hydrogen production from the Ni-Pt-washcoated ammonia-fuelled
microchannel reactor was mathematically simulated in a three-dimensional (3D) CFD model
implemented via Comsol Multiphysics™. The objective was to obtain an understanding of
reaction-coupled transport phenomena as well as a fundamental explanation of the observed
microchannel reactor performance. The transport processes and reactor performance were
elucidated in terms of velocity, temperature, and species concentration distributions, as well as
local reaction rate and NH3 conversion profiles. The baseline case was first investigated to
comprehend the behavior of the microchannel reactor, then microstructural design and
operating parameters were methodically altered around the baseline conditions to explore the
optimum values (case-study optimization).
The modelling results revealed that an optimum NH3 space velocity (GHSV) of 65.2 Nl
gcat
-1 h-1 yields 99.1% NH3 conversion and a power density of 32 kWe L-1 at the highest operating
temperature of 973 K. It was also shown that a 40-μm-thick porous washcoat was most
desirable at these conditions. Finally, a low channel hydraulic diameter (225 μm) was observed
to contribute to high NH3 conversion. Most importantly, mass transport limitations in the porouswashcoat
and gas-phase were found to be negligible as depicted by the Damköhler and Fourier
numbers, respectively. The experimental microchannel reactor produced 98.2% NH3 conversion
and a power density of 30.8 kWe L-1 when tested at the optimum operating conditions established by the model. Good agreement with experimental data was observed, so the
integrated experimental-modeling approach used here may well provide an incisive step toward
the efficient design of ammonia-fuelled microchannel reformers.
In Chapter 5, the prospect of producing H2 via ammonia (NH3) decomposition was
evaluated in an experimental stand-alone microchannel reactor wash-coated with a commercial
Cs-promoted Ru/Al2O3 catalyst (ACTA Hypermec 10010). The reactor performance was
investigated under atmospheric pressure as a function of reaction temperature (723–873 K) and
gas-hourly-space-velocity (65.2–326.1 Nl gcat
-1 h-1). Ammonia conversion of 99.8% was
demonstrated at 326.1 Nl gcat
-1 h-1 and 873 K. The H2 produced at this operating condition was
sufficient to yield an estimated fuel cell power output of 60 We and power density of 164 kWe L-1.
Overall, the Ru-based microchannel reactor outperformed other NH3 microstructured reformers
reported in literature including the Ni-based system used in Chapter 3. Furthermore, the
microchannel reactor showed a superior performance against a fixed-bed tubular microreactor
with the same Ru-based catalyst. Overall, the high H2 throughput exhibited may promote
widespread use of the Ru-based micro-reaction system in high-power applications.
Four peer-reviewed journal publications and six conference publications resulted from
this work. / PhD (Chemical Engineering), North-West University, Potchefstroom Campus, 2015
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:nwu/oai:dspace.nwu.ac.za:10394/14487 |
| Date | January 2015 |
| Creators | Chiuta, Steven |
| Source Sets | South African National ETD Portal |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0029 seconds