Rooftop Solar PV will be a game changer by Dirk de Vos

The solar PV (photovoltaic) industry is still small but it’s growing at an exponential rate with 159MW already installed up from just 35MW a year before. As the price of PV panels continues to fall and the price of electricity pushes ever upward, more people will look to rooftop solar PV as an alternative.

On the face of it, there is nothing not to like. More renewable energy means a reduced impact on the environment. There is a catch though. Our electricity supply systems are unprepared for the impact of small scale electricity generation and unless South Africa’s electricity distributors, (the municipalities) and Eskom, where it supplies directly, make the necessary adjustments, we could face a situation where everyone loses – Eskom, the municipalities, consumers, the environment, the economy and, indeed, issues of social justice.

The idea of going completely off-grid is an attractive one but in a modern urbanised context, it is not feasible. One might be able to put in a large solar array with an enormous battery but cloudy days often come together.

An analysis by Mike Rycroft ably demonstrates the point. There are seasonal differences too. The electricity a PV system produces in summer is significantly different to what it produces in winter. Our consumption profiles are just the reverse – we consume more electricity in winter. The result is that most rooftop PV installations are grid tied and there are different schemes where one can feed excess PV generated electricity back into the grid. The City of Cape Town is piloting a separate tariff where rooftop PV installations are subject to a monthly fixed charge and they buy back excess electricity, and Nersa, the energy regulator, is working on its own regulations.

Historically, the electric system has been planned only for one-way flow of electricity and, at the relatively low voltages at the distribution level, no outside generation. As such, there are limits to how much the energy the local distribution network can absorb. Reversing the power flow creates new challenges around voltage variations, overloading circuits and maintenance. These problems become more acute when renewable energy is introduced due to their unpredictable generation characteristics. Moreover, solar PV generates maximum output around midday and not during the morning and evening peaks. The PV installations in a particular suburb using the same circuit will also peak (and drop off) at the same times. A cloud passing over will affect all the installed PV panels at the same time.

The consequence is that if the amount of solar PV exceeds a certain point, say 15% of the maximum load of a particular circuit, it can cause havoc with voltage and power quality problems for everyone. Given the generation profile of solar PV, the maximum contribution of residential rooftop PV connected to the grid could not exceed 5% of total electricity generated. Better than nothing – but not by much. The problem is that even at this level, there are real impacts on the financial viability of local electricity distribution systems.

For the most part, commercial and residential electricity pricing is bundled into a per-kWh energy rate. You pay on a per-kWh basis. Even without inclined block tariffs (where you pay more per kWh, the more you consume), this pricing model places a smaller real cost-share on low consumption (poorer) customers. In recent times, this pricing scheme has worked well for our cities’ finances. As Eskom’s wholesale tariffs have shot up, the real margin that cities have been able to make from distributing this higher cost electricity has gone way up as well. The windfall, via Eskom’s tariffs, is entirely unconnected to any increased direct costs incurred by the cities in distributing Eskom-supplied electricity.

Already though, the system is falling apart. Much higher electricity prices have driven down consumption and the theft of electricity is now widespread. Johannesburg’s “nontechnical” electricity losses are especially large. Increased margins for more expensive electricity have been counteracted by fewer sales.

The particular problem that rooftop PV presents is that it is the best customers, the wealthy, high consumption customers (say those consuming 700kWh/month or more) who are deploying it first. This creates a problem not dissimilar to the way our income tax system works – the few top earners are responsible for the bulk of all taxes paid. For the electricity system then, solar PV has the same impact as a large tax avoidance scheme used by higher earners. The only difference is that electricity works as a tax on consumption so the “scheme” can’t be closed down.

Albert Hirschman, the renowned development economist, addressed the problem of the wealthy being able to opt out in his bookExit, Voice and Loyalty: Responses to Decline in Firms, Organisations and States. The wealthy, he says, have two choices when confronted with something they don’t like. They can complain, vote (voice) or they can leave and get the product/service somewhere else. He suggested that monopolies of all types might not improve their service to prevent the wealthy from leaving. The problem of bad service then rests on those who can’t leave the system, the poor. He argued that this problem applies to a range of services – a failing education system, a dysfunctional public transport system (Hirschmann developed the idea after being subjected to a horrendous train trip in Nigeria – he noted those who could leave were using cars instead). It applies to an electricity system. The key then is to have everyone largely remaining in the system so that it can be fixed for the benefit of everyone, even those who have the option of leaving it.

Rooftop solar PV is subject to a first come, first served problem which means that people who can afford to install solar do so first. If you don’t own a home, rent or don’t have the income, you can’t participate. Latecomers who might be able to afford an installation at some point in the future may well find that they can’t connect their systems to the grid because, where they live, it is already congested (at the low, low 15% of peak).

As described, the existence of solar PV creates engineering challenges on a local circuit level that, in time, will require additional resources to manage. These residential customers also are selling electricity into the grid when Eskom’s supply to the municipality is relatively cheap and coming back at peak times, when Eskom’s supply is more expensive. This reverses the existing subsidy system by shifting costs of the whole system back onto those without solar PV installations.

If far greater penetration of rooftop solar PV is inevitable, we should prepare for it and make it work for the benefit of all. First, distributors should embrace the opportunity and they could secure a far penetration of solar PV if it was carefully considered on a system basis. The unplanned first come, first served approach inherently limits how far grid tied solar PV can go. Planning for far more solar PV has another benefit. The steps required for making our electricity distributions accommodate high levels of distributed PV also addresses the serious problems that they confront today.

Like Eskom itself, our cities’ and towns’ electricity distribution systems have suffered from years of under-investment. Previous efforts to address these shortcomings, such as the establishment of Regional Electricity Distributors, came to nought. In the meantime, huge surpluses from electricity revenues have been diverted elsewhere. Any loadshedding that we experience now is no longer because of Eskom but because of the state of the local grid.

More residential and commercial Solar PV inevitably results in decreased purchase of electricity from Eskom and increases the complexity of the distribution system. It can’t be avoided. For distributors, the first step would be a detailed cost of supply analysis. How much does it cost to supply electricity to different categories of customers? For the most part, despite regulatory requirements, municipal electricity departments are not ring-fenced off from other municipal functions. A credible cost of supply analysis requires a lot of dynamic data: about Eskom’s tariffs, different customer consumption patterns and the condition of the physical infrastructure itself.

Too little of this type of data is collected notwithstanding the technology revolution in information and communications technology (to cheaply obtain, transmit real time data and to store all of it) and big data analytics (to analyse all of it). Widespread metering is needed and it can be done cheaply if done correctly.

Getting to grips with the actual cost of supply would lead to better tariff design. If distributors began to see themselves as infrastructure maintenance entities funded exclusively off the connection charge, they would be neutral as to the energy costs, whether supplied by Eskom or through distributed solar PV in the system. Once again, this depends on the implementation of time of use metering. Better data means better decisions and better regulation of rooftop PV systems.

For example, thousands of smaller 500w-1kW solar PV systems with unobstructed north facing aspects evenly distributed over a wide area would dramatically increase the total amount of solar PV possible compared to just a few large 5kW systems in our leafy suburbs. One could think of other schemes similar to the Parkhurst community broadband initiative where a suburb comes together to place PV panels on an optimal site for the benefit of the community. Large building owners, including those who are not thinking about rooftop PV, could be encouraged to rent their rooftops or other space for someone with a creative aggregation business model to install solar PV for nearby customers. What this does is to keep more of the valuable customers in the system, for the benefit of all.

There are wider national level implications as well. Our traditional top-bottom planning for electricity, as the Integrated Resource Planning (IRP) processes does, can’t work when confronted by bottom-up solar PV generation popping up all over. In fact, the very idea of having a single vertically integrated monopoly providing the whole country with baseload and peak load electricity as well as having the responsibility to maintain a synchronised system and maintaining acceptable power factors everywhere http://www.kwsaving.co.uk/Business/pfc/pfc-simple.htm is increasingly absurd.

The release of the new IRP has been postponed several times. Squaring a huge nuclear build programme into the plan while maintaining the credibility of previously published IRPs can’t be an easy task. We await another document though – the Gas Utilisation Plan which is expected to set out just how Independent Power Producers will get to build 3,000MW of gas-fired generators. For proponents of a cleaner energy system, gas is interesting. Besides being half as carbon intense as coal, it’s the most flexible source of energy around and it pairs nicely with intermittent renewable energy.

The wise thing to do is to make our cities the designated off-takers of gas-fired electricity. The economics of gas-fired electricity makes it ideal to address the daily demand peaks and use the rotating mass of their spinning turbines to keep the electricity synchronised where the load is. Eskom, with multiple problems of its own, can then focus on supplying the baseload as cheaply as possible.

Let’s restate the problem: The solar PV and associated battery (storage) revolution is being driven by technology developments (generation and storage) over which nobody has any real control. Prices for these are only going to drop. We can only decide whether we will accommodate this and make this work for all of us or ignore/resist it and see our valuable electricity infrastructure abandoned by those who can. The question is whether our politicians and regulators have the appetite to make the necessary changes. DM

Photo by ZME Science.