This blog piece is based on a recent workshop that was part of a joint project by Bridge to India, Prayas Energy Group and IIT Mumbai (National Center for PV Research and Education, NCPRE). Our research results will be published in March 2014. At the event, representatives from state and private sector utilities, government institutions and academia as well as Indian and international private sector companies including solar system installers and inverter manufacturers met to discuss the effects of distributed, infirm power generation on the grid and of fluctuating grids on PV systems. Below are some hypotheses that emerged and that will be tested further over the coming months. As always, we highly appreciate feedback.
- Detailed standards and regulations have been developed in Germany and the US to deal with safety concerns
- Even at high penetration levels, few challenges are purely technical. Most are commercial, relating to the economics of storage, balancing, spinning, etc
- Indian regulators and utilities have insufficient data to fully understand the impact of distributed power generation on the grid
At the workshop, we looked at both the effects of a PV system on the grid and of the grid on the PV system. We assumed net-metering or feeding back into the grid (which is arguably required to achieve significant numbers of distributed PV systems). We looked at safety concerns (especially islanding) and stability concerns (both with respect to the grid and with respect to the PV system). The overall mood at the meeting was that the question is timely – as there are plenty of net-metering regulations on the anvil – and that the challenges are quite manageable. The issue is more to clearly identify the challenges and develop structured guidelines to deal with them.
One of the key concerns of utilities and regulators is the safety and within it the main issue is unintentional islanding. This seems to be a non-issue, however. All quality inverters will automatically shut off, if the grid is down. Regulations from more experienced international markets like Germany and the US can be used to specify technical requirements in India in order to ensure that all new entrants in the inverter market adhere to the same standards. Intentional islanding, to allow local ‘mini-grids’ to continue to operate if the main grid fails, would require a further thought through the set of safety regulations, including communication with the load dispatch centers and utilities.
With respect to grid stability, equally, the inverter can be specified to shut off as soon as the grid frequency is outside of a specified boundary, this protecting the grid from excessive solar power. However, the details need to be clearly spelt out: What are the frequencies at which the inverter shuts off? Should all inverters shut off at the same levels, or should they be phased out at different levels, perhaps based on system sizes or location? Perhaps the tolerable frequency range in India can be larger than in e.g. Germany? What are protocols, standards and certification processes that can be put in place to ensure that systems have the right functionality right from the start? While the issue is still negligible at current penetration levels, it is easier and cheaper to specify these now than to retrofit later. At the same time, it should be avoided to burden smaller systems with excessive standards-related costs that make them unviable. There was a strong sense that distributed PV can actually help to stabilize the distribution grids.
Challenges might arise at the transformer level, if there is a lot of distributed capacity in a certain locality. If it is necessary to limit grid-interactive PV installations, installations would have to be effectively monitored, which could in itself be a difficulty. However, transformers typically have a loading capacity of 110% of the rated load and have safety devices to cut off if that is exceeded. Also, it might be unrealistic to expect solar to exceed the load. Wherever the power demand is high with respect to available space/rooftops, e.g. wherever cooling is used, this is unlikely. The issue might arise in rural areas. However, here the economics of feeding solar into the grid without FiTs are not strong as power prices are heavily subsidized. So this might not be an issue after all.
Another challenge is to manage sudden, sharp drops in PV generation (e.g. if there is a sudden cloud cover). These issues will only arise at high penetration levels, but could be relevant in specific places already quite soon. Gandhinagar in Gujarat, for instance, is planning to install a total of 10 MW of solar capacity (8 MW on rooftops) with a base load power requirement of only 40 MW (25%). This will be an interesting test case for understanding the impact of a high penetration of solar generation on the grid. Inversely, the opinion at the workshop was that voltage fluctuations in the grid and outages of the grid have no significant impact on the functioning or lifetime of PV systems.
The larger grid-management questions relating to stability, such as balancing the grid, providing storage and spinning reserves for times when renewables are not available, pricing the solar power fed back into the grid and pricing distribution and transmission services for this power are not so much technical issues than commercial ones. They relate to the proper incentivization. However, if they are not addressed, they lead to technical challenges of managing the grid. There is little clarity on what would be a dangerous limit of having infirm, unbalanced power in the grid. According to an older NREL study, 20% should be unproblematic. In Germany, the grid has functioned well with penetration rates of up to 60%. Similarly, ‘smart grids’ that allow for a very active communication between different generation and consumption points are not so much a technical necessity as an economic opportunity.
The thrust of the research should be on clearly outlining and structuring the various challenges that need to be addressed and to give Indian and international case studies and data showing how they have been tackled already to increase the comfort level of regulators and utilities. In addition, there should be standardized (modular) requirements for different system types and sizes, including e.g. protocols for grid interconnection points and detailed inverter spec sheets. From the utility and regulatory side, processes need to be standardized to provide fast turnaround times for any permits needed in order to not stall the market.
Tobias Engelmeier is the Managing Director at BRIDGE TO INDIA.