A techno-economic feasibility study on the use of distributed concentrating solar power generation in Johannesburg
Date
2010-09-21
Authors
Bode, Christiaan Cesar
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This study provides an evaluation of Concentrating Solar Power (CSP) technologies
and investigates the feasibility of distributed power generation in urban areas of
Johannesburg. The University of the Witwatersrand (Wits) is used as a case study
with energy security and climate change mitigation being the main motivators.
The objective of the study was to investigate the potential of CSP integration in
urban areas, specifically investigating Johannesburg’s solar resource. This is done by
assessing the performance and financial characteristics of a variety of technologies
in order to identify certain systems that may have the potential for deployment.
To aid the comparison of the technologies, CSP performance and cost data which
were taken from multiple sources, were adjusted giving it local, present day
assumptions. A technology screening process resulted in the conception of twelve
alternative design configurations, each with a reference capacity of 120 kW(e).
Hourly energy modelling was undertaken for Wits University’s West Campus for
each of the twelve alternatives. Three configurations were further investigated and
are listed below; each with a design capacity of 480 kW(e).
1. Compound Linear Fresnel Receiver (CLFR) field with an Organic Rankine
Cycle (ORC).
2. Compound Linear Fresnel Receiver field with an Organic Rankine Cycle that
integrates storage for timed dispatch.
3. Compound Linear Fresnel Receiver field with an Organic Rankine Cycle that
integrates hybridisation with natural gas.
Levelised electricity costs (LEC) of the systems were used as the basis for financial
comparison. Real LECs, for the three configurations above, range between
R4.31/kWh(e) (CLFR, ORC) and R3.18/kWh(e) (CLFR, ORC with hybridisation).
v
With the energy modelling of the hourly direct normal irradiation (DNI) input into
the CSP systems, Wits University’s West Campus Electricity bill was recalculated.
The addition of the solar energy input resulted in certain savings and a new LEC that
is Wits-specific. These LECs ranged between R3.98/kWh(e) (CLFR, ORC) and
R2.77/kWh(e) (CLFR, ORC with hybridisation). A third LEC was calculated that
integrates a CSP feed-in tariff (REFIT) of R2.05/kWh. At the time of writing, a CSP
REFIT of R2.10/kWh was released which favours the analysis.
The analysis of the 480 kW(e) systems resulted in total plant areas of between
10350 m2 (CLFR, ORC,) and 15270 m2 (CLFR, ORC, with storage). With plant
modulation, these plants can be placed on vacant land, above parking lots or on top
of buildings which would also provide shading.
The values obtained for the average yearly insolation was 1781 kWh/m2 based on
TMY2 data. Johannesburg has a very intermittent source of DNI solar energy. The
summer months in Johannesburg yield a higher peak DNI, whereas the winter
months provide a more consistent average. This is due to the high amount of cloud
cover experienced in summer. With this insolation, CSP electric generation is
possible however, compared to the other locations, it is not ideal. Also, because of its
intermittency is has been advised that certain applications such as HVAC and
process heat and steam requirements be pursued.
From the results, it can be concluded that power production costs through small
scale CSP systems are still higher than with conventional fossil fuel options,
however several options that may favour implementation were recognised. Through
the analysis it was found that if the CSP generated electricity is valued at the market
price ( CSP REFIT), the payback time of such systems can be decreased from 73 to
12 years (CLFR, ORC with storage). Further, due to the scale of the plants analysed,
the exploitation of high efficiencies and economies-of-scale of plants with power
levels above 50 MW(e), is not possible. With the introduction of these technologies
vi
at lower power levels, cost savings through the incorporation of other design
options (such as waste heat utilisation) should be pursued.
It was recognised that South Africa in general has one of the greatest solar resources
in the world and should therefore be technology leaders and pioneers in CSP
technology. With greater emphasis being placed on the need for renewable energy
systems, it is imperative that South Africa develops its skills and a knowledge base
that will work at making the implementation of renewable energy, and in particular
CSP generation, a reality. Technologies identified that should be pursued for
distributed generation include Linear Fresnel collectors that are easy to
manufacture and don’t involve complicated receiver systems. There is also scope for
developing thermal storage technologies in order to make generation more reliable.