Integrating Renewable Energy Resources: a microgrid case study of a Dutch drink water treatment plant
MetadataShow full item record
In the Netherlands 1.2 billion cubic meters of drinking water are produced each year, which uses about 480 GWh of electricity per year as an energy intensive industrial process. While there are some examples of renewable based microgrids around the world that illustrate the benefits of producing and consuming renewable electricity onsite, there has been little research on the potential of a renewable-based hybrid system at a much larger scale and in a Dutch context where wind speeds are higher, solar irradiation more moderate, and where subsidy schemes for renewable energy investment exist. Thus, this research aimed to identify the technical and economic potentials of integrating solar photovoltaic (PV) and wind power with demand response into a grid-connected versus stand-alone microgrid at a Dutch drink water treatment plant. It also aimed to investigate the economic implications for flexible demand in the water pumping process with demand response (DR) and the extent of reliance on grid-imports, storage, and back up generation. Lastly, it aimed to evaluate the economic potential of the microgrid system in 5 years if investment costs would further decrease. Method. Waternet’s drink water treatment plant in Nieuwegein (DWP-NWG) was used as a case study to model a microgrid integrating solar PV and wind power with demand response by researching its electricity demand, the production potential for renewable electricity (RE) supply, and the flexible demand potential in the pumping process. These were then used as inputs to model a variety of grid-connected and stand-alone microgrid cases using HOMER modelling software under the current financial support scheme and in 2018. The main technical potentials were evaluated based on renewable electricity production potential, fraction of onsite electricity demand supply for renewable electricity, reliance on grid imports, back up diesel production, and storage throughput. Economic potentials were evaluated based on the system’s levelized cost of electricity (COE), net present value (NPV), and discounted payback period (PBP). Results. The DWP-NWG has the potential to supply 70-96% of its annual electricity demand with 17.8-25.6 GWh/year of locally produced solar PV and wind power combined with demand response. The most profitable system configuration is a 5.6MWp Solar PV – 8MW wind (4 turbines) combined with demand response, yet without storage. This microgrid system can be 88% self-sufficient on renewable electricity, with a COE of 0.020 €/kWh versus 0.082 €/kWh for the current situation without renewables, an NPV of €12.7 million, and a payback period of 7.3 years. The least cost stand-alone system is a 5.6MWp solar PV – 8MW wind – 1.3MW diesel generation – 2MW cell stack/45MWh electrolyte flow battery storage with demand response, which has a COE of 0.094 €/kWh and an NPV of €-7 million. If electricity prices rise, grid-connected potentials increase by 16-55%. • Demand response: Comparing a 4 wind turbine system with demand response to one without demand response, the former has an NPV of €11.8 million and the latter €11.2 million. This shows that demand response adds about €600,000 of value over a 25 year project lifetime. Annually, shifting 29% of normal annual demand to be supplied during renewable electricity production earns an additional €59,000 per year in avoided transport and energy taxes. Larger capacity microgrid systems, which can pump slightly more water above normal demand from the excess RE produced, earn up to €71,000 per year. The maximum capacity RE system with DR and without batteries can utilize 12% of the additional flexibility provided by the water storage buffer at the Dunes; however, in order to utilize the maximum 15% flexibility from the buffer at the Dunes, adding battery storage is required in order to defer the use of excess electricity when the strongest pumps are not being used. This is due to the limitations of the current pump installations and transport network which prevent significantly more water to be pumped in one time step. • Grid imports, back up generation, storage: Even at these high potentials, both grid-connected and stand-alone cases still rely on 1-4.5 GWh of electricity per year imported from the grid in grid-connected cases or supplied by back-up diesel generators in stand-alone microgrid cases. This is due to the intermittency of solar PV and wind power production. Grid-connected microgrids don’t necessarily need storage, yet stand-alone microgrid cases require 2.2-3.2 GWh of annual storage throughput. • 2018 Potentials: If investment costs further decrease and no tax incentives of subsidies exist for RE, the most profitable microgrid case configurations remains the 5.6 MWP solar PV – 8 MW wind (4 turbines) – DR without battery storage configuration with a COE of 0.030 €/kWh, an NPV of €10.7 million, and a discounted PBP of 11 years. The least cost stand-alone microgrid has a COE of 0.118 €/kWh, and an NPV of €-11 million. If the EIA & SDE+ financial support is still available in 5 years, NPV potentials are 40% higher. Conclusion/Discussion: The technical results show that a significant portion of electricity demand of an industrial sized water treatment plant can be met with a few wind turbines and large PV capacity. They also confirm that a solar-wind combination is optimal over a wind-only configuration. The economic results indicate that stand alone microgrids are not cost-effective, while grid-connected microgrids are very profitable. This is because directly consuming renewable electricity onsite avoids the majority of electricity costs, sell-back of excess electricity is possible, and because the impact of investment costs has decreased since the SDE+ subsidy decreases wind investment costs by at least 50% and solar PV investment costs have drastically dropped by over 40% over the last year. Moreover, based on the sensitivity analysis of electricity prices, the profitability of these systems is even underestimated if electricity prices rise in the future. The results also indicate that while the technical potential to shift electricity demand with demand response is significant, the economic advantages are relatively small since the margin between buy-in and sell-back electricity prices is marginal for wholesale electricity consumers. Nonetheless, the investment for demand response could be paid back in less than 2 years and will become more profitable as electricity prices rise and sell-back rates decrease. Moreover, a 100% renewable system would require extremely large battery storage, which is not currently cost effective. Ultimately, even at the low wholesale electricity and sell-back price for large industrial electricity consumers, grid-connection and the ability to trade excess electricity is extremely important for the cost-effectiveness of microgrid system. Lastly, although PV costs have already dropped significantly making the current economic potentials from grid-connected solar-wind microgrid configurations very high, waiting to invest within 5 years can be even more profitable as long as the EIA and SDE+ support remain.