Early Tuesday morning, news of lower than 2¢/kWh bids from Saudi Arabia’s first round bids on 300MW of 9.5GW solar power projects started popping up. Seven of eight bids were below 3¢/kWh, and two of eight bids broke a prior record low. Questions have arisen as to the sustainability of the bids outside of these unique circumstances.
The Abu Dhabi Future Energy Company (MASDAR) and renewable energy group EDF Energies Nouvelles placed a bid for 25 years at 6.69736 halas/kWh – equivalent to approximately 1.79¢/kWh. A second group, ACWA Power, bid ~2.34¢/kWh. 100hal=1 ryal=~0.27USD
The 1.79¢/kWh is the first time we’ve had a bid breach 2¢ – I took a gander that this barrier would be breached back in February when the Saudi’s announced that they’d offer ‘motivating terms’.
These bids are not yet final, “All of the bids opened today will now undergo stringent technical, financial and legal evaluation,” the Head of the Energy Ministry’s Renewable Energy Project Development Office, Turki al-Shehri, told Reuters in a statement. The bids, shown in the following image, were officially via Twitter by the Saudi Arabia National Renewable Energy Program.
The project’s pricing has taken many by surprise.
Aymen Grira suggested that, ‘The big volume, the long-term commitment, the attractive financing and the sovereign warranty have made the Middle East a land for profitable opportunities in the solar market.’
Jenny Chase of Bloomberg suggested pricing is partially so low to get companies into the door for the larger 9.5GW project. Also that there is a decent chance some form of price escalator is involved.
“What most people suspect, including us, is that there is some sort of subsidy, either direct or indirect, within the PPA that makes this price possible,” said Benjamin Attia, an analyst in global solar markets at GTM Research. “Or they’re going to lose money on it. It’s very unlikely in my mind that Masdar is making money on an unsubsidized 1.79¢ (USD) bid.”
Browning Rockwell – spoke on the fundamental involvement of the state as parties winning bidding groups: “ACWA Power is closely aligned with the Saudi Arabian government and royal family’s interests from its origin, and has received equity investment from the Saudi Public Investment Fund. Masdar is part of Mubadala, a state-owned company in Abu Dhabi.”
I actually first heard of the bid via Ian Bollard on LinkedIN early in the day. Since this bid has come out, and the consternation on the viability of the bid leading to several proclamations on its near illegitimacy has followed it, Ian thought it appropriate to comment and pat the back of those involved in the project that had put in the work and earned this pace-setting bid.
The main driver, in my opinion, is that the state actors bidding are able to gain access to cash (their own banks) at some of the lowest prices on the planet. Something is behind the two record-setting bids being made by state actors – while the majority of the bids stayed in more ‘sane’ regions between 2.6-3.4¢/kWh. Two other factors – newer, more efficient solar hard and customer specific finance tools – probably had some effect on these record prices.
These state actors have multiple variables affecting their financial models that pure corporate entities cannot account for, such as: future revenue from selling more oil abroad vs burning for electricity, increased employment for middle class workers within an evolving Saudi Arabia, knowledge of changes in future retail electricity rates (only de facto ruling actors have such cross over power), and of course Saudi Arabia requiring local content. Nation-state techniques 101 – giving cheap capital to build local facilities seems to be a great logic – political and economic victories. These particular countries are in a unique financial position of often looking for intelligent ways to deploy their large cash resources as a result of oil sales – their sovereign wealth funds and cash is invested across global investments and stock exchanges. Capital heavy investments like solar power are a good way of keeping the money at home, feeding the locals with good work, who increase the local velocity of money – pumping tax coffers.
Having cheap money – and being your own risk analysts (aforementioned de facto power have broad ownership) – give the rest of us a view of what solar might cost with ‘free’ money and high levels of interparty trust.
The main driver of cost, in a large solar project in Saudi Arabia, will be that of hardware. There are many techniques of paying manufacturers. For instance, you could pay them their project ‘profits’ up front so they can invest in the manufacturing lines before production begins (always great to have sold the first 12 months of production in advance), and then paying off hardware as part of revenue sharing over the 25-year electricity contract. Solar panel manufacturers are in a cash-poor situation, while being within a solar manufacturing hardware buying cycle having to upgrade to the newest mono-PERC panel line technology, why not some triangulation with the cash-heavy group and the machine makers? Maybe some of those manufacturers are building local facilities as part of the local content requirement?
The other driver of hardware – advancing technology – is also lowering pricing and that part of this project is much more applicable to broader global game. The first is increasing individual panel efficiency. With the price of near tomorrow’s 19-20% MONO-PERC higher efficiency solar panels coming in line with today’s standard 15-16% efficiency panels we are going to see significant labor cost savings. Jinko just announced a solar cell over 22% – that means the leading low cost manufacturers are going to break even bigger efficiency records soon.
For math’s sake, when installing endless racks of 300MW of solar panels with panels that are 18% higher in efficiency (16.2% at 315W versus 19.5% at 379W), you will install 161,172 fewer solar panels. This will lower the effective cost per watt of labor, from say, 15¢/W to 12¢/W. For a project like this – probably costing in the 70-80¢/W range – 3¢/W on labor represents up front savings of ~4% of the total project’s costs. Other cost savings from efficiency arise also – fewer racking components needed, less copper cabling, shorter machine rental periods, less land needed, few interconnection transformers, etc. Savings here of just a few cents a watt, as seen above with labor, can have a significant affect on project cash flow. Imagine the labor savings using a bifacial solar panel that is in the 22-24%+ range (or even that 21% panel from Jink mentioned above) – these desert areas have highly reflective surfaces and might be suitable for this type of tech.
Shifting projects in this class downward 6¢/W lower’s initial cash outlay from, say $242M at 80¢/W all in, to $222M at 74¢/W – $20M cold hard out of pocket cash saved.
Most finance teams will still question the already low $222M number though. A rough calculation says this project will generate 552GWh/year – that will equal $9.8M in revenue at 1.79¢/W. If the loan were given at an absurdly low-interest rate of 1% on a $222M loan – you’d have to pay back $10M/year – real close to the $9.8M in annual revenue from the plant. If this project is financed at 0% – standard Islamic Financing forbids interest on loans – then the project can squeeze through on payments of $8.8M. The way the $/kWh are noted – many digits out from a whole number (6.69736) – could suggest the bidders worked backwards from an annual payment: Figure out exact annual payment to meet finance terms, divide by projected production = price we have to charge per unit. Of course, projects producing this much annually need more exact pricing (extra digits) as the effects of rounding become very large very quickly.
One commenter noted that the electricity production estimates used might be highly conservative, so I thought to show the full variables used in my production calculations – capacity factor of 21%, based upon an estimation here and a comment from a smart person elsewhere. Total Annual Output = 300MW times Capacity Factor 21% times 24 hours in a day times 365 days in a year equals 551,880,000 kWh/year. Moving from 21% capacity factor to 22% means an extra $470k/year. Increasing capacity factors could come from higher efficiency, better cleaning techniques, fewer losses in broader system design, etc. Our from afar guesstimates of energy productions are a bit dicey, when we see how such a small amount of production can affect back side revenue generation.
You can clearly see how the project’s revenues press tightly against what is possible even with the world’s lowest interest rates, using the newest technology and cutting edge financial tools. There is also clear indication that small estimation changes in product efficiency, and electricity production might change some serious project variables.
With projects of this size, slight changes in how cash is distributed to involved parties can change moods. And, with a heavily involved deal making process, it is possible to make many groups happy with different finance techniques. These techniques will be a challenge to replicate outside of these banking standards, interested related state/private markets and wealthy countries – however – they will help Saudi Arabia maximize its investment when installing 9.5GW of solar power.
While the rest of the world might not be able to build programs exactly like Saudi Arabia’s – there are many other tools to get amazing low pricing globally. Mexico and Chile found a way under 3¢ and the USA has, multiple times, puts itself within the 3.1-3.4¢/kWh range – and this is with taxes on solar panels. In Mexico and Chile’s case, it is world-beating sunlight and conducive construction atmospheres – the USA offers tax credits and low finance costs as well. There is no ‘free market’ for energy – there is a complex, intertwined global market of ideas carving up greater than 100GW of solar power in 2017. And maybe it’s acceptable, and expected, that the terms of deals in Saudi Arabia won’t translate globally.
A cool perspective on the bid event – live webfeed image via twitter showing the bid coming through just before 7.40 AM EST –
Header image is ‘free for public use, its reproduction and adaptation is authorized provided the following source is acknowledged: “SolarGIS © 2014 GeoModel Solar.”‘