A Case for On-Site Renewable Generation

In my last article, I pointed out the relevance of the Grid as a potential blocker for our climate change goals (in Mexico and elsewhere). This time, I want to call your attention to an enabler: on-site power generation, a model that allows us to not only continue pushing for the energy transition, but, also, to diversify our way into energy security.

The term is self-explanatory but there are enough derivatives of the concept to justify a few clarifying lines. For instance, on-site generation can be done at different scales, from rooftop panels to several MWs of capacity or even GWs for industrial facilities (large scale is less common simply because a lot of land is required, and these necessarily adjacent areas were usually not planned for). Also, on-site generation can be interconnected to the Grid or can be completely isolated. The latter variant is oblivious to the Grid but requires self-sufficiency in return. Considering that most residential-level generation is photovoltaic and even though it can be complemented with increasingly affordable storage units, full self-sufficiency today is hardly viable. Still, some military facilities have been working on isolated systems because they understand them as less vulnerable (the conflict in Ukraine is proving them right). Hybrid systems that combine clean(er) generation technologies with complementary production patterns are also gaining traction in other industrial applications.

Now, let’s go through some of the implications of a more prominent role of on-site generation. First is the green aspect: the more each "generating site," be it a regular home or an industrial complex, produces renewable energy, the less utilities and/or power companies need to generate under a centralized logic. To the extent that large-scale generation is fossil, on-site generation reduces emissions. However, if the "sites" are not fully independent, baseload generation, the kind that is versatile and fully predictable (and for the most part fossil-fired), is still needed. Thus, when it comes to producing energy, the on-site model is a simple zero-sum game with a minimum baseload restriction, where part of the investment responsibility and inherent risks are transferred to consumers. In return, consumers access a viable business case as  a way to achieve their own environmental goals and an opportunity to hedge against fuel prices and improve the service reliability they get from their current suppliers. Conversely, the utilities’ role in the system would need to evolve too, shifting  from a day-to-day struggle to keep a massive generation portfolio operational, toward a focus on better managing cleaner and more predictable generation, such as nuclear or the use of long duration storage at scale, retiring unnecessary antiques and properly revamping, operating and maintaining a reduced “gray” baseload capacity.

What about the impact on-site generation has on the network? The need for capillarity at the distribution level is clear but the network also needs to be smarter, capable of precisely measuring more injection points and predicting their behavior to make good use of this injection while avoiding imbalances. This requires some investment but at different orders of magnitude and implementation speed than redundancies needed under the centralized model. On the other hand, with the correct use of storage and the right network code, on-site generation can contribute to the network’s reliability.

Granted, the case is not straightforward as it calls for systemwide modeling for a number of trade-offs, but I would argue that at the highest level of abstraction, the trends are easier to read: while on-site systems are becoming more affordable and consumers more willing to generate their own power,large-scale infrastructure is not getting any easier or cheaper to obtain, operate and maintain, particularly “thermos”. Also, let’s not forget that most of the world has been fully penetrated by distribution networks, but between 1-2% of the population still lacks access to power. Self-supply is the quickest way to solve this. 

Unfortunately, it is easy to imagine how incumbents would feel compelled to resist more participation of on-site generation. After all, they are losing clients and there are more assets to consider in the network. Let’s have a look on the contrasting cases of Mexico and Brazil: 

As a result of these clearly opposing policies, Brazil enjoys spectacular growth in on-site generation across consumption levels (only distributed generation has +13GW installed, with increasing levels of centralized generation under the self-supply scheme, which is probably facing limitations or even full incentive cancellation by means of new regulation). Meanwhile in Mexico, industrial self-supply is stagnant and residential continues to grow but at a much lower scale than Brazil’s (+2GW in DG and a self-supply model under regulatory attack). But unlike the southern giant’s, the Mexican system is currently short of clean and reliable sources of energy, causing many enterprises and investments to be delayed or plainly redirected elsewhere. 

Finally, my intention is not to say that centralized generation does not have a key role for transition purposes anymore, far from it. But instead, to recognize that we need to explore all possible ways to decarbonize our energy models. On-site generation has proven to be an effective model and, as such, we should be putting our heads together to further propel it at both the residential and industrial levels. Given that the Brazilian and Mexican leaders apparently are in such a close relationship and in the hunt for arguments to demonstrate leadership, will they listen to each other’s ideas on energy policy?