Geographical optimization of variable renewable energy capacity in China using modern portfolio theory
In: Applied Energy. Applied Science Publishers: London. ISSN 0306-2619; e-ISSN 1872-9118, more
|   |
| Author keywords |
Variable renewable electricity; Efficient frontier; Portfolio; Geographical smoothing; Capacity factor; Geographical potentials |
| Authors | | Top |
- Hu, J.
- Harmsen, R.
- Crijns-Graus, W.
- Worrell, E.
|
|
|
| Abstract |
The large-scale deployment of wind and solar, which are variable renewable electricity (VRE) technologies, is indispensable to decarbonise China’s power sector. However, variability in VRE outputs poses challenges in power system operation in terms of increased demand for backup and reserve capacity. These challenges can be effectively mitigated by “geographical smoothing”, because spreading VRE deployment over a large area largely reduces the variability associated with the collective output of VRE. Based on meteorological reanalysis data, this study characterised the return and volatility (i.e. mean and standard deviation of hourly capacity factor) per VRE asset in China at a high-resolution grid cell level. This enabled to identify the efficient frontier of optimal VRE portfolios that capture the geographical smoothing effect for China’s future power system, using modern portfolio theory. The portfolio volatility is minimized for each attainable return. We analysed key statistics of optimal portfolios, including technology shares, levelized cost of electricity and capacity factor at-risk values. Our results show complementarity between wind and solar in China, reflected in more optimal return-volatility performance of wind & solar portfolios, as compared to wind-only and solar-only portfolios. In addition, our results show that portfolios with unconstrained technology shares perform much better in return-volatility performance than portfolios with constrained technology shares. This suggests existing scenarios in literature with pre-defined shares of different VRE technologies might be sub-optimal. This study also shows that for optimal wind & solar portfolios a “firm” non-zero minimum portfolio capacity factor (1.4–5.5%) can exist with 100% availability. |
|
