Modelling the impacts of high solar and battery uptake on the electricity market

About the modelling

3 reports were commissioned to examine the potential effects of high rooftop solar uptake on wholesale market outcomes, distribution networks, and household energy costs.

The reports considered Rewiring Aotearoa’s forecast of around 28 GW of non-utility (distributed) solar generation by 2050-2055. The modelling contrasted a moderate uptake base case (around 3.5GW by 2055) with 2 uptake trajectories: moderate and fast uptake rates to achieve Rewiring’s target of 28GW for rooftop solar adoption (including some with attached battery storage) by 2055.

The results provide useful indicators of benefits and challenges that may arise if rooftop solar levels increase significantly. The reports do not tell us an optimal uptake pathway for New Zealand.

Modelling the impacts on the wholesale electricity market

Jacobs’ wholesale modelling exercise predicted significant effects on the wholesale market: the large volume of low-cost non-utility solar could lead to very low summer prices and more volatile winter prices. This model suggests this effect could last until 2035, after which time grid-based generation investment would be required to continue serving the expected ever-increasing demand. The model also predicted significant changes in hydro storage: with lake levels being much higher in summer but drawn down more heavily in winter.

This modelling used the same demand profile for each scenario. Future modelling could examine a higher electrification demand, and/or a lower solar uptake level than 28 GW.

Modelling the impacts on low voltage distribution network investment requirements

ANSA’s distribution investment-impacts modelling exercise looked at the effects of high levels of solar and battery uptake on costs relating to distribution network investments, and consumer benefits derived from battery uptake.

The report used the same solar growth rates as the wholesale modelling, and also explored scenario combinations involving early and late electrification, along with different scenarios of after diversity maximum demand (ADMD) increases of 0 kW, 1 kW, and 2 kW per ICP (consumer).

The report concluded the capex requirements on networks in almost all scenarios are likely to be substantial if solar rooftop exports and electrification peak demand are unmanaged. However, if solar rooftop exports are highly managed (e.g. via use of household-scale batteries, hot water energy storage, local EV charging, and some curtailments on peak generation days), then the increase in capex requirements from solar rooftop exports could be substantially mitigated.

Modelling the impacts on household energy consumption bills

Sapere’s research assessed the potential effects of the modelled wholesale and network impacts (described above) on household energy costs. These costs comprise electricity bills, costs relating to rooftop solar and battery installations, and gas costs where these are present. This stage also explored how different tariff structures, which change incentives around the timing of EV charging and solar export, could affect electricity bills.

This modelling strongly supports the importance of a smart energy system in New Zealand. This will require not only the introduction of smart meters and control systems but also appropriate incentives and tariffs to ensure consumers (or their panels, batteries and EVs automatically) can shift demand and injection to the ‘right’ time periods – so that customers benefit via lower bills driven directly by changing their time of use; and indirectly via avoided network investment costs. Without such ‘smartness’, there would be a greater risk of mismanaged, excessive or stranded investments, resulting in higher costs to consumers.

Addendum to the Modelling the impacts on household energy consumption bills report

A separate addendum to the Modelling household energy consumption bill report, prepared with Rewiring’s input on the modelling outputs, discusses some of the learnings from this work regarding suitability of input modelling approaches, assumptions used and discusses what may need to be considered for future work.

Download the reports

3 separate reports were commissioned and are linked below: