-----------------------------------------------------------------------------------------------------------------------
ECOS2003,
16th International Conference on Efficiency, Costs, Optimization,
Simulation and Environmental Impact of Energy Systems,
Copenhagen, Denmark
June 30 - July 2, 2003
http://www.ecos2003.dtu.dk/
----------------------------------------------------------------------------------------------------------------------------------------------------
Report from: Federico Zenith (Ph.D. student working on Hydrogen
Systems Technology, supervised by S. Skogestad)
ECOS 2003 was held in Copehagen from June 30 to July 2. The
conference had
most of the time 5 parallel sessions and occasionally some plenary
sessions. Most European countries were represented, with Denmark being of
course the one with most representatives. It was maybe surprising that so
few Americans came (11), fewer than Italians.
From NTNU, there was Signe Kjelstrup, who will organise ECOS2005
that will be held in Trondheim
(ECOS2004 will be in Mexico). I met also Gelein de Koeijer from the
Statoil research center at Rotvoll.
The proceedings will be available to the institute in paper form as soon
as I return to Trondheim, but are available in electronic form as a pdf
file at the following address:
users/zenith/ECOS2003.pdf
Remark that this file is over 80 megabytes and includes the equivalent of
three books.
On the Sunday 29 the organising university, DTU, offered a tour
of the Avedoere power plant, that delivers about 850 MW of
electricity and
about 900 MW in district heating. The plant had a pleasing architecture
and the environment looked very little polluted in its surrounding, even
though it is primarily a coal plant; maybe it was because the second gas
turbine was down for maintenance. The plant is as said primarily based on
coal, but features biomass, natural gas and wind as alternative sources
of
power. A few pictures can be seen at this address:
users/zenith/avedoere/
Sulphur obtained from coal combustion is recovered as CaSO4; we were as
lucky as to come at a time when the second, newer turbine was stopped for
maintenance, and we could see it unmounted (see link to pictures
above)
Afterwards, there was a Matlab workshop held by none other than
Cleve
Moler, "father" of Matlab, that by the way has the
strongest American
accent I've ever heard. As there were many beginners, the workshop (just
2
hours) was quite general, but there were some interesting demonstrations
with fractal theory ("fern" function).
Conference, plenary session:
The conference was started by Septimus van der Linden,
chairman in
ASME-IGCI, with a plenary lecture on "Development trends and the
role of
gas turbines in the future 30 years". He drew an interesting
comparison
with the first plane of the Wright brothers and its development in 30
years, and defined the main factors for gas-turbine development: alloys,
cooling technology, thermal coatings, fuel flexibility, composite
materials etc. An efficiency goal of 75% for 2015 was proposed. The
alternative of nuclear-powered hydrogen generation was discussed,
stressing the importance of not using too much cooling water as today. It
was also stressed how distributed generation will take over larger
plants,
which means that the Brayton cycle will have to be kept simple. A final
goal was to shorten the time of maturation of new technology from the
current 6-8 years to just 2.
Next was Paul-Frederik Bach from Eltra, that spoke about
"Simulation and
Optimisation in a liberalised stochastic market". He pointed
out that
electricity and heat demands do not overlap when discussing CHP
generation, and also that the woes experienced in Norway because of last
year's draught have had noticeable effects on Denmark and Sweden as well
through the NordPool market. Models of energy systems have difficulties
implementing all the various sources currently in use; especially
start-stop costs and wind power seem to be difficult. Modelling can be
used to identify bottlenecks and must be updated for distributed
generation, today over 50% of local generation in western Denmark. Wind
power has shown itself being especially unpredictable, with errors of
over
a few hundreds MW for the danish market. NordPool is a day-ahead
market,
while disturbances in wind power have to be eliminated continuously, so
simply selling electricity to NordPool is not a solution.
Future models will have to implement game theory, the fact that not all
players behave rationally, wheather forecasting, power-system
architecture
and analysis; a new market simulator for nordic countries and northern
Germany, dubbed MARS, was briefy described.
Jens Rostrup-Nielsen from Haldor-Topsoee spoke about
"Sustainable energy
through chemical process technology. Prospects for a hydrogen
economy?".
His presentation started with serious problems with computer crashes and
I
will let you guess which OS he was using.
He compared prices for production of hydrogen, that favour clearly
natural
gas (1-5 EUR/GJ), while NG with CO2 removal is more costly (8-10 EUR/GJ).
The most expensive source is solar, 40-50 EUR/GJ, but breakthroughs are
awaited in photoelectrochemical cells that will produce hydrogen without
passing through electricity and electrolysis. A description of the
reforming process followed for the uninitiated.
Flaring of NG was determined to be an important source of CO2: in Asia,
an
impressive 15% of total NG production is simply flared. The
Fischer-Tropsch was proposed as a solution to stop this waste, since the
process has been significantly improved since the days of the
Stukas.
Mr. Nielsen posed the question whether it would be more sensible to use
NG
to make diesel, and save money to research real solutions instead of
"patches" as hydrogen from NG. He made a good point in that
there are no
resources of Pt available to fit all vehicles with PEM fuel
cells.
(Technical sessions begin)
"Optimum operation strategies for thermal power plants in the
deregulated
electricity market", by Carraretto and Lazzaretto. Basically
dynamic
optimisation with some case studies, but nothing that is not in the
syllabus of SIE1035 Energiplanlegging.
"Methods for calculating CHP electricity", by Folke.
Quite a confusing
presentation mostly on the definition of CHP in Europe and the need of a
EU directive to define it.
"Power-market dynamics: the statistical mechanics of transaction-based
control", of Chassin. He underlined the different nature of centralised,
radial systems and decentralised, distributed systems, the latter of which
are considerably more complex and exhibit different behaviour. He outlined
a market-based control system based on a simgle theoretical unit, the
general user/producer (or ATM, abstract transitive machine).
"Integrated model-based optimisation of a power-production system", of
Tommy Moelbak. He started out with a provocative "We don't need modelling
and control, we need optimisation in an integrated sense". He analysed the
various control levels in a whole power-production company, and came to
the conclusion that what a company needs is optimisation, which _may_
require advanced simulation and control, but these are not an objective
themselves.
"Production of lime, hydrogen, and methanol by the thermo-neutral combined
calcination of limestone with partial oxidation of methane or carbon" by
Halmann and Steinfeld started a series of titles that looked more like
abstracts. The presented paper was well described by its title, and
indicated a payback of less than one year.
"Exploring new production methods of hydrogen/natural gas blends for
mixing into the natural gas network of the Netherlands", presented by
Leslie Zachariah, was about reducing CO2 emissions by "greening" the net
gas; the Netherlands have a very extensive NG network. Drawbacks are
mixing of high-value hydrogen with lower-value natural gas, and loss of
energy in the conversion. A not-so-efficient, S-tolerant catalyst could be
used as there are no such things as purity requirements; H2 content is
limited to about 17% before utilities have to be updated to e.g. new
burners.
The economical drive is that consumers will pay the same money per cubic
meter for a lighter gas, thereby paying more for their energy.
Oh well...
"Syngas purification by means of membranes for hydrogen based power
cycles: preliminary analysis" by Lombardi, Corti, Fiaschi and Tapinassi
applied a ceramic membrane to a gas turbine cycle.
Srinavasa Murthy from India presented two papers in a row: "Screening of
metal hydrides for engineering applicaitons" and "Combined heat
transformation, heat pumping and refrigeration using metal hydrides". In
the first lecture he introduced a number of properties of metal hydrides
and their desired properties, introduced concepts as the isteresis factor
in MH and presented the results of his experiments on thousands of
different materials. In the second lecture, he applied MH to a maybe not
so common area, i.e. heat pumping and transformation.
The second day opened with a plenary session with Cleve Moler about the
evolution of Matlab (that, as others do, was obstinately written in all
caps as MATLAB). It started out with a description of the early years of
Matlab, the founding of the Mathworks, and the live discovery by Moler
that the original "Matlab classic" (the one before 1.0) does not run under
windows XP. It was interesting to hear that Matlab toolboxes are actually
released under a "half-open source" policy (what this means in practice I
don't know exactly). He made an application example with the case of long
jumper Bob Beamon, that set a long-lasting world record in the long jump
at the Mexico City olimpics. He claimed, after showing a model and a
simulation's results, that he was not helped by the marginal contribution
of reduced air drag due to the height of Mexico City, but rather he
focused much more on the running part of the jump, which had been long
neglected by the jumpers, who often arrived to the jumping line with too
little kinetic energy.
He stated that Matlab's objective regarding performance is to be "as fast
as C", and showed the effect of the Dynamic Type Analysis and the
Just-in-time Compiler embedded in Matlab; each accelerates performance
roughly by a factor of 10. He predicted that future versions of Matlab
will keep as much as possible detail hidden to the user, keeping "the math
inside".
(back to technical sessions)
"Finite-time thermodynamic optimisation of mitochondrial chemistry and
fuel cells" by Andresen and Essex showed an interesting parallel between
our hydrogen fuel cells and the biological "power plants" of our cells,
drawing a paragon between "homo sapiens" and "machina automobile" (one of
the rare istances where the use of latin is still appropriate). The
similarity went on with the reactions ADP=>ATP and H2=>H2O, but the rest
of the paper was mostly about biological arguments.
"A scheme of computation, analysis, design and optimisation of solar
Stirling engines" by Michael Feidt (same as above) et.al. was another
reminder of why one should learn languages before delivering a lecture.
It was mostly about the prediction of performance of solar-dish Stirling
engines and how to optimise them.
"How to avoid overcooking: optimal yield of B from the consecutive
chemical reaction A<=>B<=>C", presented by Peter Salamon, was mostly a
theoretical work on a problem that has no complete analytical solutions
(for the case of all reactions being reversible). It involved optimal
control over a given time span, the manipulated variable being
temperature. Most disappointing was the fact that Mr. Salomon did not
implement any conditions on the use of the manipulated variable
(temperature), which resulted in the solutions being presented as "hold
infinite temperature until t1, then cool down".
"Multi-objective optimisation of remote-community energy systems with
consideration of noise disturbance", by Pelet, Leyland and Favrat, could
have been a very interesting paper: it was about optimising the power
delivery to a small oasis in the desert, 150 km from the nearest sign of
civilisation. The object function was a function of noise level and cost.
In a process equivalent to varying the weights on these two terms, it was
found that solar power would have been competitive if the fuel cost had
been doubled (e.g. because of a CO2 tax). A number of case studies were
presented to show how the optimal distance from the oasis varied depending
on the importance given to noise levels.
However, again, the lecture was spoiled by the poor English sported by the
speaker. It took me a while (and I did study some French) to understand
what "seh-oh-doh" was. (CO2)
"A relaxation-based heuristics for the design of cost-effective energy
conversion systems", by Ahadi-Oskui et.al. concentrated on optimisation of
cogeneration plants. The paper presentation had a sharp
optimisation-theory, pure-mathematics approach, and it was a bit difficult
to follow due to the quantity of information that was crammed in the
presentation.
"A method for the economical optimisation of indirect counter-current heat
exchangers coupled in series" by Finnish professor Pertti Sarkomaa focused
on the economic optimisation of n indirect heat exchangers between two
streams, i.e. how many and how large they should be. It was a bit unclear
why one would need more than one at all, and professor Sarkomaa pointed
out "safety reasons" as for example avoidind certain fluids to contact
with each other. Among the interesting things, the definition of "economic
efficiency" of an exchanger.
"Minimising the entropy production rate in two heat exchangers and a
reactor" by Audun Roesjorde from NTNU Chemistry institute, with the
co-authoring of Signe Kjelstrup, focused on entropy production in a
reactor that converted propene to propylene, with an undesired reaction
(cracking) to be avoided. A few simplifying assumptions were made, among
which the probably invalidating one of no pressure drop across the heat
exchangers, thus neglecting an important source of entropy: all phases are
gaseous, and in order to obtain good heat-transfer coefficients one has to
pay in terms of pressure drop.
"Overall efficiency analysis of hydropower plants using optimisation
techniques", by Wulff Hirano from Brazil, showed that Norway is not the
only country depending substantially on hydro power: Brazil makes almost
80% of its power in hydro plants, located mostly in the southern part of
the country. The hydro plant of Itaipu alone, with its 12600 MW, could
cover a fifth of Italy's consumption. In 2001 shortages were experiences,
but situation is now back to normal. The optimisation was carried out
essentially to maximise the efficiency, and among the conclusions there is
that using two different turbines will broaden the high-efficiency area.
"A method for the economical optimisation of the design temperatures and
connecting flows of a heat pump system" by Pertti Sarkomaa was essentially
an application of concepts developed in the previous presentation to
another area, optimising the economic efficiency of heat pumps by changing
their sizing parameters.
"Perspective for the integration of biomass in combined-cycle power
plants" by Franco, Giannini and Casarosa of the university of Pisa
scratched all previous records of boredom. The lecture dragged on for one
hour instead of 20 minutes, with many complex flow diagrams, quite bad
English, long formulas and almost all "things not to do" you can find in
any given book on how to present papers. Students from DTU that were
"drafted" to take pictures of the event later told me that they did not
take pictures of the audience because they did not want to portray an
audience that was about to fall asleep.
I mention this presentation only because it really taught me something -
be brief, clear, concise, say what's important and then stop.
(third day)
Yukitaka Kato from the university of Tokyo presented a paper on "Thermally
regenerative fuel reformer for a zero-emission vehicle". This was a very
well executed presentation, possibly the best in the whole conference
among the ones I attended, and the idea is as well very interesting: CaO
would be used to generate the heat needed in the on-board fuel reformer to
obtain hydrogen with the exothermal reaction CaO+CO2=>CaCO3; this CaCO3
will later be recovered and purified thermally of CO2, that will be
stocked, and reutilised again as CaO. This system will therefore give a
0-emission driving capability. A few experiments were also presented, that
demonstrated how the CO level was kept low (<1%), though not yet low
enough to be acceptable to a PEM stack.
The weight of such a system would be very little (about 7 kg), especially
compared with equivalent equipment with metal hydrides for hydrogen
storage (100 kg).
"Evaluation of different configurations of PEMFC-based powertrain
systems", presented by Andreassi, university of Rome-Tor Vergata, took
data for a typical engine usage cycle (so-called "European cycle") and
calculated from speed and acceleration the voltages and intensities needed
from a fuel cell stack; comparing with a cycle with many stop-&-go, i.e.
traffic jam (the "NYC cycle"), and different degrees of hybridisation with
batteries, the conclusion was that battery hybridisation with fuel cells
is generally good and lowers H2 consumption. The optimal degree of
hybridisation is instead not always clearly determined.
"Economic assessment of home cogeneration system using PEFC taking into
account fluctuating characteristics of hot-water demand pattern",
presented by Takeyoshi Kato, focused on japanese conditions and analysed a
soon-to-be-commercial fuel-cell unit produced by Mitsubishi, which will
convert natural gas into eleectricity and heat. Its performance will be
very dependent on gas prices.
"Graphical exergy analysis of a fuel cell" by professor Ishida was mostly
a theoretical work that was quite difficult to follow in the presentation,
but it may be an interesting alternative way to look at exergy analysis.
"Development of a model for thermoeconomic design and operation
optimisation of a PEM fuel-cell system", by Frangopoulos and Nakos from
the university of Athens, was a work that optimised the present life-cycle
value of a fuel-cell stack, and came to the conclusion that design for a
lifetime of 6-7 years is optimal.
"Modelling a PEM fuel-cell natural-gas reformer", by Nielsen, was another
study of reforming applied to fuel cells; among the points, real reformers
are too complicated to be implemented in small applications and have to be
simplified. The algorithm and a case study for methane were presented.
"Simulation study of a PEM fuel-cell system with autothermal reforming",
by Sibel Ozdogan from Marmara university in Istanbul, was mostly another
reforming study developed in HYSYS.
"A novel approach to the thermoeconomic analysis of fuel-cell and
gas-turbine hybrid-power systems from the fuel-cell experimental data",
presented by Antonio Valero of the university of Zaragoza, concentrated on
molten-carbonate fuel cells (not exactly the most promising), and
determined that exergy destruction ratios are lower for hybrid systems; he
presented a prediction of fuel-cell costs in the next years, considered
that fuel cells can half CO2 emissions and have neglectable NOx and SO2
emissions, and claimed that MCFC will be competitive with advanced
gas-turbine combined cycles by 2015.
"Thermoeconomic analysis of a SOFC-based 3-generation system", presented
by Panopoulos of the university of Athens, was actually not
thermoeconomics (by admittance of the speaker who did not like the title),
but had the very interesting point (comprehensibly neglected in a country
like Norway) that combined heat and power is not so attractive in Greece
and southern countries in general, since there is generally low demand for
heat at all. An absorption-refrigeration cycle was then taken into account
for a CHCP (combined heat, chilling and power) system, that would be more
attractive as there is much more pressing demand for chilling in summer
months.
"Preliminary studies on an ambient-pressure MCFC/steam-turbine combined
plant based on thermo-economic considerations", presented by Piero Lunghi
of the university of Perugia, had the very interesting point that gas
turbines, though often seen as the prime alternative for bottoming cycles
to recover high-value heat in high-temperature fuel cells, are not so
attractive for MCFC, which have temperature ranges of 500-600 celsius for
their exhaust. The calculations of efficiencies of fuel-cell/gas-turbine
combined systems are optimised and lose a lot at part load, while steam
turbines are presented as more flexible. A model in ASPEN+ was presented.
This is a very interesting study, though MCFC are not that important,
because the same ideas can be applied to new-generation SOFC which will
work at lower temperatures and have very attractive properties.
Final plenary sessions:
"Compenting models for promotion of wind power in Europe", by Niels Meyer,
presented as the man "guilty" of the massive danish investments in wind
power, was a very interesting speech about wind power ("almost
competitive", according to the speaker), but also a lot about the current
energy policies of many countries. Meyer was especially critical of the
lack of environmental concern, attention of supply security and the
dependency on fossil fuels, that could start new wars, and the allusion
to the latest Iraq war was more than clear. He claimed that an inherently
short-sighted market cannot be entrusted to find long-term solutions.
The price of wind power has been estimated to 0.035 EUR/kWh, which is
competitive with governmental subsidies, which have recently been declared
legal by the EU.
He observed that wind power cannot "compete" in the common meaning, since
it cannot be manipulated.
"Photovoltaics at the threshold of commercial breakthrough" by Eirik Sauar
from the norwegian Renewable Energy Corporation presented the growth of
solar as 5 years ahead of schedule and running. Multi-layer solar panels
with over 70% efficiency have been built for space applications (ISS),
while theoretical calculations show the limit to be at 92%. The most
important markets are Japan (largest) and Europe (fastest growing),
especially Germany.
"The introduction of venture capital into the renewable and distributed
energy market sector", by Even Bakke, focused on the importance of
distributed generation and of developing countries, and again the date of
2015 was indicated as the breakthrough of fuel cells against combustion.
The alternative energy sector has been estimated to grow between 20 and
30% a year. It was claimed that it follows closely the NASDAQ index.
(the end)
This archive was generated by hypermail 2.1.3 : Fri Sep 05 2003 - 16:42:10 CEST