Assessment of menu of contracts' incentive properties under different designs of sharing factor functions

Abstract

Incentive regulation has become challenging due to the increased information asymmetries and uncertainties driven by the diffusion of distributed energy resources (DER) and smart grid technologies. A combination of a Reference Network Model (RNM) and a menu of profit-sharing contracts has been used in this thesis to address these problems in electricity distribution. RNM is a tool which can be used by regulator to estimate the investment cost needed by distribution system operators (DSOs) and thus, the problem of information asymmetries between regulator and DSOs can be mitigated. Menu of contracts, as defined in this thesis, is a regulatory scheme with the revenue determined ex-ante and reviewed ex-post, albeit based on some pre-defined rules. It ensures DSO to receive the greatest reward when the forecast investment cost coincides with the true expenditure in that regulatory period. Therefore, the use of menu of contracts can encourage truth-telling and hence avoid strategic behaviour of DSOs. Consequently, this regulatory mechanism has drawn some attention from regulators as a mean to tackle the aforementioned increasing uncertainties. The sharing factor is a key parameter that is needed to build a menu of contracts. It determines the strength of the incentive given to the DSO, i.e. how much network companies would benefit from cost reductions or how much they would be penalized for an increase in costs as compared to allowances. In a conventional profit-sharing contract, the sharing factor is a constant value. However, under the menu regulation considered in this thesis, this parameter is obtained as a function of the ex-ante investment cost estimation submitted by the DSO. Four sharing factor functions have been designed: two functions with a different rate of change when the DSO/regulator ratio increases (increasing and decreasing rate of change respectively) and two functions with an asymmetric sharing factor dependent on whether the difference between actual expenditure and allowed revenue is positive or negative. These sharing factor functions are applied in the menus of contracts, together with the network expansion cost estimated by RNM at different levels of PV penetration to assume different realizations for regulator’s forecast cost, DSO’s forecast cost and actual expenditure in menu of contracts. Sharing factor functions with different rate of change across DSO/regulator ratio provide the regulator with higher flexibility in setting incentive strength of the menu. The analyses show that, when the regulator has a high level of confidence with the benchmark cost, a sharing factor function with increasing rate of change when DSO/regulator ratio increases can be particularly useful, and vice versa. While for a sharing factor function which varies with actual expenditure, it is possible for the regulator to reward outperformance and penalize underperformance at a different rate, especially when one outcome is more desirable than the other. In case the regulator wishes to deter overspending over cost-saving, a sharing factor function which increases with the actual expenditure can be used. On the other hand, when the regulator wishes to have the investment projects completed rather than to avoid overspending, a sharing factor function which decreases with actual expenditure can be used. All in all, these sharing factor functional forms can be used to achieve different requirements desired in the remuneration scheme. Careful tuning of the parameters used in these functions makes the menu of contracts more flexible in term of having different rate of change of sharing factor with DSO’s estimated cost and with actual expenditure.