Feedheat-integrated storage for load-following nuclear

Abstract

The conversion of primary energy into electricity involves a series of energy movements and transformations. This ‘conversion pathway’ can potentially be exploited to integrate energy storage in a more efficient and cost-effective manner, when compared to conventional electricity-in-electricity-out storage systems. The cost-benefits occur since the conversion hardware (which already exists for electricity generation) can also be used for the storage processes. As such, generation-integrated energy storage (GIES) systems have recently been of interest, with proposals for wind, solar, and nuclear energy. In future low-carbon energy systems, it is likely that a proportion of electricity generation will come from nuclear steam plant (or biofuel steam plant). During the conversion of nuclear fuel to electricity, the steam plant produces large heat flows, which presents an opportunity for integration with thermal storage. Not only would a nuclear GIES system provide grid flexibility, but it would also reduce the proportion of intermittent generators on the grid; thereby reducing the total system storage requirements and the level of required renewable over-generation. Our previous work has shown that ‘sensible’ thermal storage integrated with the feedwater heating system is highly effective. The feedwater system uses the condensing turbine bleed flows (at their various pressures and temperatures) to lift the temperature of the feedwater by 200C. This ‘staircase’ heat-transfer process (as it looks on a T-Q diagram) exploits the low temperature difference between the condensing bleeds and the sensible feedwater, and provides good thermal matching with a sensible storage medium. Round-trip storage efficiencies are greater than 80% and the net work output can be flexed by 10% (100MWe for large nuclear). This integrated-storage process operates by varying the amount of feedwater in the feedheating system: During charge -- the amount of feedwater is increased above its nominal value, requiring more steam is bled from the turbines. The excess of heated feedwater is then transferred to storage. By over-bleeding the turbines, the net work output is reduced. During discharge -- a fraction of feedwater is diverted from the feedheating system and heated via the storage system. Since the feedwater system is now underloaded, the bleed flows are reduced and net work output increases.

Throughout charge, discharge, and nominal operation, the heat input from the steam generator remains constant, allowing the reactor to operate at a 100% capacity factor. This is crucial, as it allows the most expensive part of the plant (the reactor) to operate constantly at maximum output, while simultaneously allowing the electricity output to vary and ‘load-follow’. (In the UK, this system is effectively removing the financial constraints which currently restrict nuclear to operate as baseload.) During our previous analysis the off-design performance of the heat exchangers and turbines was considered, to ensure integration of storage did not adversely impact system performance. It was determined that departures from design conditions have only a small impact, since the change in magnitude of turbomachinery and feedheat losses are inversely proportional to each other. This paper builds upon our previous work by optimising the design of the steam plant for a new-build system which maximises: (1) the throughput efficiency and (2) the % flex of the whole system. During this process the marginal increase in plant cost will be considered (e.g., to what extent does the marginal cost of increasing the heat-exchange area of feedheaters justify the increase in rational storage efficiency?). Specific optimisation parameters will include heat-exchanger area, bleed pressures, and characteristics of the thermal storage system. The three main system efficiencies (storage, nominal plant, and throughput) will then be calculated for a range of operating scenarios and varying charge/discharge durations. Different integration options, such as integration around only the high-pressure and low-pressure turbines, are also considered. Additionally, the cogeneration of heat is explored, and methods to calculate effective COP are discussed.