1. Introduction
In recent times, solar thermal power plants (STPPs) have
attracted interest as a large scale, commercially viable way to
generate electricity [1]. In an STPP, the heat transfer fluid (HTF) and
the working fluid play an important role as the carriers of energy
from the collector/receiver to the turbine. This is commonly done in
two stages for a plant operating with a Rankine cycle. The HTF
(e.g. synthetic oil, molten salt, etc.) first collects the energy from the
incident solar radiation. This energy is then passed on to the
working fluid (water/steam) which carries it to the steam turbine.
The main disadvantage of such two-fluid systems is that the
maximum operating temperature of the HTF is limited by the fluid
stability concerns (e.g. approximately 400 C for the synthetic oil),
thus resulting in a low turbine inlet temperature and consequently
a low cycle efficiency.
Application of direct steam generation (DSG) in STPPs presents
the prospect of improving the overall plant efficiency, while
simultaneously decreasing the cost of electricity generation [2]. The
pressurized steam is generated directly in the receiver and transported
to the steam turbine. The advantages of DSG include a
higher live steam temperature and the use of one fluid as both the
HTF and the working fluid, possibly resulting in a simplified operation.
The main disadvantage of using DSG for STPPs is that it
requires a very complex storage system for uninterrupted plant
operation [3]. The motivation behind the current study is that the
exergy losses during a heat transfer process can be reduced by
using a suitable multi-component working fluid which can evaporate
or condense at a varying temperature, contrary to the constant
evaporating or condensing temperature for a pure substance
[4]. One such multi-component working fluid is the ammoniawater
zeotropic mixture, as used in a Kalina cycle (KC). There have
been discussions regarding the feasibility of using ammonia-water
mixtures at high temperatures due to the nitridation effect
resulting in corrosion of the equipment. However, the use of an
ammonia-water mixture as the working fluid at high temperature
has been successfully demonstrated in Canoga Park with turbine
inlet conditions of 515 C and 110 bar [5]. Moreover, a patent by
Kalina [6] claims the stability of ammonia-water mixtures along
with prevention of nitridation for plant operation preferably up to
2000 F (1093 C) for temperature and 10,000 psia (689.5 bar) for
pressure using suitable additives. It should be noted that the term
direct steam generation is used here for both water and ammoniawater
mixtures.
There were proposals to incorporate the KC for waste heat recovery
plants, geothermal power plants or solar energy driven
power plants. Such plants operate with low or medium range
temperatures at the turbine inlet. Bombarda et al. [7] presented a
thermodynamic comparison between the KC and an organic
Rankine cycle (ORC) for heat recovery from diesel engines. They
concluded that although the obtained electrical power outputs are
nearly equal, the KC requires a much higher turbine inlet pressure
to attain the same, thereby making it unjustified for such use. Singh
and Kaushik [8] presented energy and exergy analysis and optimisation
of a KC coupled with a coal-fired steam power plant for
exhaust heat recovery. They found out that at a turbine inlet
pressure of 40 bar, an ammonia mass fraction of 0.8 gives the
maximum cycle efficiency and that the highest exergy destruction
occurs in the evaporator. Campos Rodríguez et al. [9] presented an
exergetic and economic comparison between a KC and an ORC for a
low temperature geothermal power plant. They found that the KC
produces 18 % more power than the ORC with 37 % less mass flow
rate. In addition, the KC had 17.8 % lower levelized electricity costs
than the ORC. Wang et al. [10] presented a parametric analysis and
optimisation of a KC driven by solar energy. They found that the net
power output and the system efficiency are less sensitive to the
turbine inlet temperature under given conditions and that there
exists an optimal turbine inlet pressure which results in maximum
net power output. Coskun et al. [11] presented a comparison between
different power cycles for a medium temperature
geothermal resource. They found that the KC and the double flash
cycle provided the least levelized cost of electricity and hence the
lowest payback periods.
With regards to using the KC with high turbine inlet temperatures,
Ibrahim and Kovach [12] studied the effect of varying the
ammonia mass fraction and the separator temperature on the cycle
efficiency for a Kalina bottoming cycle using gas turbine exhaust as
the heat source. The KC turbine inlet conditions were 482 C and
59.6 bar. The authors found that the KC is 10e20 % more efficient
than the Rankine cycle with the same boundary conditions. Nag
and Gupta [13] performed an exergy analysis of a KC with gas
turbine exhaust as the heat source with a turbine inlet temperature
between 475 C and 525 C, and a turbine inlet pressure of 100 bar.
They concluded that the important parameters affecting the cycle
efficiency are the turbine inlet temperature, composition and the
separator temperature. Dejfors et al. [14] presented an analysis of
using ammonia-water power cycles for direct fired cogeneration
plants with a maximum temperature of 540 C. They concluded
that for a cogeneration configuration, the Rankine cycle performs
better than the KC whereas for the conventional condensing power
application, the performance of the KC is better. Knudsen et al. [15]
presented the results from the simulation and exergy analysis of a
KC for an STPP having a turbine inlet temperature of 550 C when
the heat input is from a solar receiver, and 480 C when the heat
input is from a molten-salt storage system. The authors varied the
heat input to the cycle so as to maintain the turbine inlet conditions
while assuming the same mass flow rate for all the cases. Modi et al.
[16] presented a comparison between a Rankine cycle and an
ammonia-water cycle for STPPs with a turbine inlet temperature of
450 C. The cycle energy efficiency and the storage size requirement
were used as the comparison parameters. With regards to the
analysis of central receiver STPPs, Xu et al. [17] presented the energy
and exergy analysis of a central receiver STPP operating with a
Rankine cycle. They concluded that the efficiency of the plant can
be increased by focussing on reducing the losses in the receiver and
by using advanced power cycles.
A recent review of research on the KC by Zhang et al. [18]
highlights the use of KC for various applications like bottoming
cycle, low temperature geothermal, industrial waste, etc. In the
review [18], and to the authors’ knowledge, there were no studies
on using the KC for high temperature STPPs with DSG. The purposes
of the current study are to assess the potential benefits of using a KC
for a central receiver STPP with DSG using exergy analysis, analyse
the trend of the rate of exergy destruction in different components
of the plant with respect to the pressure and the ammonia mass
fraction at the turbine inlet, and compare the performance with a
simple Rankine cycle (SRC). To attain these objectives, the KC was
modelled and optimised for maximum work output for the
assumed boundary conditions and analysed for operation when the
heat input was only from the solar receiver, or when the primary
source of heat input was a two-tank molten-salt storage system.
The ammonia mass fraction is defined here as the mass of ammonia
in the ammonia-water mixture to the total mass of the mixture. The
paper is structured as follows: Section 2 presents the assumptions
and the modelling procedure, Section 3 presents the results from
the exergy analysis and the operation from molten-salt storage
system, Section 4 discusses the results and Section 5 concludes the
paper.