Closing the ‘digital divide’ in areas currently without reliable access to power can not only change lives, but can also create huge development and business opportunities. But this will only be possible through improving energy access, a new report has found.
A collaboration between Bloomberg New Energy Finance and social media giant Facebook, the Powering Last-Mile Connectivity report was undertaken for the Telecom Infra Project’s Power and Connectivity Working Group, co-chaired by Facebook and Spanish broadband provider Telefonica, which aims to ensure affordable and reliable energy access for key connectivity use cases beginning with rural areas.
The report focused on the nearly four billion people who are currently not connected to the internet. It found that, even with growing electrification in the world’s least developed countries, nearly 700 million people will still lack electricity in 2030.
A change in course is needed, the report warned, in order to avoid continued limitations on electrification due to infrastructure costs, low demand in rural areas and, in some places, a lack of political will.
But according to the report, a course change is eminently possible as the costs of distributed power systems fall and new business models arise. Decentralized Energy spoke with Itamar Orlandi (pictured), Head of Frontier Power at BNEF, to discuss the findings.
Q: What kinds of issues or awarenesses gave rise to the paper?
A: There is a nexus which has always been there, but in the paper we tried to articulate it with better examples and more clearly. Energy is a very significant enabler of better connectivity, and connectivity can unlock benefits in a vast variety of sectors and aspects of life. If the connectivity sector is interested in reaching those people without internet access, then energy needs to be a component of that solution.
The report was designed to highlight the fact that questions around energy access and interconnectivity access are much more closely related than is typically recognized. This happens both on the customer side, so we talk about issues with customers who struggle to keep their phones charged and the extraordinarily high cost they pay relative to data consumption and device ownership, as well as on the infrastructure side, where telecoms struggle to get power to existing towers.
Our estimate is that the industry currently spends $3.8bn on diesel fuel to power those sites. As you can guess, hybrid solutions – diesel/energy storage and diesel/solar/storage – can do that on a cheaper basis. We are starting to see uptake in this, but if you want to expand connectivity, the cost of that last tower is a unit cost. If you can reduce it further, you remove a barrier to increasing connectivity access.
The report also talks about possible alternative technologies from the connectivity side that consume less power and could therefore be feasible for applications in remote areas. There’s a lot of OPEX as well as O&M going on when towers are fuelled with diesel, and this is a burden on mobile network operators (MNOs) or power operators.
Q: The issue of last-mile connectivity seems to be coming to the fore in discussions of rural electrification schemes. As our focus is decentralized energy, one question we have is whether last-mile solutions are expected to allow grid power to put distributed solutions out of business?
A: There is limited evidence of that. Even given current trends, the UN goal is still out of reach for 2030. If you extend the lines of where energy access is trending in percentage terms, we’re not going to get close by 2030. We’re going to solve a fraction of the problem at the current pace.
One thing that is important to keep in mind when thinking about extending the grid is that we’ve seen tremendous improvement in energy efficiency, and this is most pronounced in very basic household items like lighting, entertainment and, to some extent, even smartphones’ power consumption and all the different use cases you can get with that. The energy needs of basic households in remote and low-density rural areas in developing economies are so low that, with conventional power consumption, utilities don’t manage to make a business case for extending the grid there. If consumption is a fraction of what it used to be 10 to 20 years ago, it’s much harder to build a business case to extend the grid into those areas, as the absolute cost of extending the grid is same. Similarly with telco towers, many of which are in areas where consumption is not significant enough to justify extending the grid.
We don’t take a normative position on whether energy access should be grid-connected or distributed. The pure cost per kWh is going to be cheaper for a long time if you’re grid-connected, and there are no incentives to disconnect in most cases. The question is going to be twofold: can you justify the overheads involved in extending the grid connection, and is the payback period short enough – and often there’s a question around that as consumption is very low in those areas.
The other question is around speed. Grid connections are centralized government or utility-driven projects that can take quite a long time, as evidenced by the fact that 100 per cent of the world is not connected. At the very small end, distributed technology is basically something you can ship in a box, and it functions much more like a retail product than an infrastructure service.
Q: For decentralized power, how important are technological solutions vs new business models – and is that balance changing in favour of the latter?
A: I think the only reason why there is fresh momentum around the last-mile question is the rapid cost reduction in, originally, PV, but over the last few years also in storage. It’s important to keep in mind that this has nothing to do with developments around R&D on last-mile reach or energy access. The energy access space’s momentum is a derivative of things happening elsewhere.
The hardest part of a lot of the opportunities outlined in the paper is on the R&D side around improving the energy efficiency of connectivity hardware, or improving product integration: taking off-the-shelf components, optimizing them for particular applications and integrating them. The focus is on improving not energy efficiency, but operational efficiency. You can have a faster rollout more cheaply by leveraging partnerships through existing distribution channels, logistical networks and optimized data use to better identify and target customers and get overheads down.
So I think that’s basically the answer: a lot of innovation will have to come through deployment and business model innovation, while relying on existing technologies where economies of scale in other sectors justify more R&D.
Q: How best to encourage these new partnerships to drive electrification? We’re interested in the implications of Facebook’s involvement as well as mobile network operators and other companies driving electrification as a means of creating new markets – will they take over from NGOs as the most significant boosters of the drive for rural power?
A: To a large extent, within all of the approaches we outline, one niche within that space, pay-as-you-go (PAYG) solar, is the one with the most traction so far. What has worked there is relatively targeted deployment of patient capital and impact capital that has helped to get the attention of more commercial investors.
I think a lot of these sectors are still at a pretty early stage, so they have applied for patient capital or some of the grand approaches we outlined that enable experimentation. Once you go one step ahead of that, you see that, for example, adding solar to powering towers is capital-intensive. That’s the big change when you go from diesel, which is low CAPEX/high OPEX, while solar is the opposite.
The question is: how can you create models that get the debt required for these kinds of technologies or deployments into the correct vehicles. PAYG is an experiment in off-the-balance-sheet vehicles, while powering telco assets is an ESCO model which is an emerging approach: players that do the O&M but provide financing service as well.
Q: Which companies are the major drivers for growing connectivity, and in which regions?
A: What we’re seeing in PAYG solar is recent actions from established companies. A few months ago we saw Fenix International taken over by Engie, and GE Ventures and Shell Technology Ventures announcing investment rounds in this space. So you do see an increased interest and activity around the sector. In PAYG there tends to be a prolonged transition and it’s still a blended approach. We’ve seen some success in getting commercially-minded investors engaged with this, and financing levels have increased.
What you start to see is recognition that, within the 1.1 billion people without energy access, one section will be commercially viable to serve and another section is going to be harder, so there will be roles for both kinds of financing approaches. This is based on a variety of drivers, but the question is where you can justify the overhead. Here factors such as purchasing power, population density and whether circumstances allow you to run a business in a place come into play.
Imagine a very low-end solar lantern, available below $5 (in some places at least). That’s one approach, and that technology can be deployed either purely through NGOs or through a market approach. In many places it’s done through a market approach by companies that will sell basic products with relatively low margins in an attempt to gain consumer confidence and, eventually, to be able to sell more valuable products to the same customers. So these are attempts to reach those customers, and even do PAYG for even smaller assets. But the incentives for commercially-minded PAYG companies will be to serve customers who can create higher margins for them, and that may be in areas where population density is higher or ability to pay is higher.