Solar power has come of age. Not only has it become one of the key alternatives to fossil fuels, it can now be deployed in a way that makes a viable business with a financial profit. This book shows industry professionals and students how to do just that.
Solarnomics describes the economics of building and operating a solar power plant today and provides a window into a future in which several technologies collaborate, and in which all participants in the electricity grid become smarter at scheduling both the supply and demand for electric power to give humanity a future that is sustainable, both environmentally and economically. The book shows how to estimate costs and revenues, how to tweak the design of a project to improve profitability, how to calculate return on investment, how to assess and deal with risk, how to raise capital, how to combine solar with batteries to make a hybrid microgrid, and how to be prepared for future developments in the evolving smart electricity grid.
Solarnomics will enable professionals in the solar industry to assess the potential profitability of a proposed solar project, and it will enable students to add an extra dimension to their understanding of sustainability.
Librarian Note: There is more than one author in the Goodreads database with this name. This profile may contain books from multiple authors of this name.
David Wright combines an Engineering PhD from Cambridge University with his current position as Professor at the University of Ottawa’s Telfer School of Management
solar power can be proftable without government incentives. It is published in research journals with the top 2–3% of citations.
Let’s say “Modules” instead of “Panels”.
harnessing solar power is not the only way to generate renewable electricity. Hydroelectric and wind are the two major alternatives.
Due to the rapid pace of improvements in solar installations, the construction cost is declining at 12% per year. Hydro power is a very different story. The technology and its installation are well established and do not have much room left for further tweaking to reduce costs. We have already dammed the vast majority of the rivers that can be dammed easily, so new projects are in more challenging locations. Consequently, the average cost of hydroelectric power installation is increasing by 4.4% per year. Wind power tells a similar story. Although it is windier offshore than onshore, installation costs are very high for offshore wind turbines and there has been relatively little deployment except in shallow waters. With just modest technological advancement, costs of onshore wind farms are declining by only 2.7% per year, much more slowly than the 12% annual reductions for solar.
Of those 260 GW of renewable capacity installed in 2020, solar had the largest share
Batteries are declining in cost at 11% per year, about the same rate as solar, but are currently expensive, eating into the proft margin on our renewable power plants.
could we make our electric vehicle battery do double duty: to power a car and also to power our home after we have driven home?
“What is the proftability of solar?” “How do we assess the proftability of solar?”
Direct current (DC): electric current that always fows in the same direction; electricity produced by a solar module is DC
Float-o-Voltaics: solar modules foating on water
Feeder line: electric power line in a city used by a distribution company to deliver power to many customers, each of which has a connection to the feeder line
Frequency sensing and regulation: ability of an electric load to sense the AC frequency of the electricity grid and respond by increasing/decreasing demand when the frequency is too high/low
Gigawatt hour: a unit of electric energy = 109 watt hours
Global horizontal irradiance (GHI): the total solar power on a horizontal surface including the diffuse light from the whole sky plus the direct beam of sunlight, taking into account the angle of incidence
HOMER: hybrid optimization model for electric renewables; commercial software for optimizing the design of a renewable energy system
Hurdle rate: the minimum IRR required for a project to go ahead
Hybrid: electricity generating system (e.g. a microgrid) that has more than one type of generator, e.g. solar, wind, tidal, and diesel
Independent system operator: a company responsible for balancing supply and demand for electric power in the public electricity grid, e.g. by operating wholesale markets for power
Irradiance: a measure of the power of sunlight in kWh/m2
Load: (1) electrical device that consumes electricity, e.g. a domestic refrigerator or a conveyor belt used to transport rock in a mine, and (2) amount of electricity consumed by such a device
Modules: also known as panels; mounted on racks, these are the building blocks of a solar installation. Each module consists of several solar cells covered with protective layers of plastic and held together in a metal frame. Solar modules typically produce 250–400 Watts of DC electricity and are 0.6–2 square meters in area
Multijunction solar cell: a solar cell consisting of multiple semiconductors on top of each other, each converting a different part of the solar spectrum into electricity
Pedestal: a vertical support for an array of solar modules that allows the modules to follow the sun by rotating about two axes: east–west and high–low tilt
Sky camera: a camera with software that monitors the passage of clouds across the sky to forecast solar electricity generation a few hours ahead
Step tariff: a tariff that increases the charge ($/kWh) for electricity when the customers’ monthly consumption exceeds a certain limit; also known as tiered pricing
Surety bond: an agreement between an insurance company and two parties, A and B, to a contract. If A fails to meet their contractual obligations, the insurer makes a payment to B. Surety bonds are widely used on solar construction projects, between a developer and a contractor, e.g. if an electrician fails to complete electrical work on time, a payment is made to the developer
Tracker: a system that swings solar modules around so that they always point at the sun; single-axis trackers follow the sun from east to west; dual axis trackers also follow the sun as it changes altitude
How much will it cost? How much revenue will it bring in? What factors are beyond our control that inevitably limit our revenues or add to our costs? In what ways can we tweak the design of our project to improve revenues or reduce costs? How much return will we get on our investment? Should the project go ahead?
Solar modules do one thing and one thing only: they use semiconductors to convert light into electricity. This is known as photovoltaics (PV): photo (light) to electricity (volts). used in 96% of today’s solar installations: crystalline silicon.
seen through the lens of proftability.
A residence may pay a fat rate price for electricity at any time of day or year. An offce building may pay a base price for electricity plus a charge that depends on the peak consumption of the building.
The sun rises in the east and sets in the west, and in the Northern Hemisphere, during the day, it passes through the southern sky, so we set up solar modules facing south. In the winter, the midday sun is low in the sky, and in the summer, it is high; the average altitude angle of the sun is 90° minus the latitude of our location. If we are installing fxed* solar modules, we tilt them at an angle approximately equal to our latitude to generate the most electricity during the course of a year. If we are near the equator, they are hardly tilted at all, facing almost straight up. In Perth, Western Australia, they should face north with a tilt of approximately 32°, and in Frankfurt, Germany, they should face south tilted at 50°.
On the roof of an offce tower, in a windy city like Chicago we might tilt them much less than the latitude (42° for Chicago), because a storm would cause undue stress to the modules themselves, the racking and possibly to the roof of the building. A fatter installation of 10° − 25° might be required by local building codes. Suppose we install them at 22° in Chicago, i.e. 20° away from optimal, we can expect a reduction in electricity output by about 6%.
Let us defne utility-scale projects to be at least 2 MegaWatts (MW) of capacity and many are in fact much larger, a few hundred MW. To get an idea of what a 2 MW system looks like, imagine 5,000 modules, each with a capacity of 400 W and a surface area of 2 square meters (m2). To set 2 MW in perspective, the largest solar farm is at Bhadla, India with 2,200 MW, not far off the largest operational nuclear power station at Kori, South Korea with 7,500 MW. Philippines has three projects of 1,200 MW each.
he sun shines from a different angle in the sky depending on the time of day and time of year. It generates the maximum electricity when it is at right angles to the modules. The more oblique the angle, the less power is generated.
If we also want to track the variations in altitude of the sun, we need a two-axis tracker.
It turns out that the extra electric power from a two-axis tracking system is usually not worth the additional cost. However, a single-axis tracker only adds 7% to the total cost and can pay for itself in extra electricity generated. In places with clear skies, this is worthwhile, but in locations near cities with hazy skies, or mist or cloud in the morning or evening, the additional electricity does not pay for the tracker. In the USA, about 70% of new utility-scale installations use singleaxis trackers.
In deserts where land is zero-cost, spacing between rows can be used to reduce shading. Where there is a land cost, we need to make a trade-off between paying for space and reduced electricity generation due to shading.
P45 graph
Soft costs include land acquisition, obtaining permits, interconnection to the electricity grid, overhead and profit for the developer. Structural and electrical balance of system includes the ground supports, racking, wiring and other electrical equipment.
We achieve considerable economies of scale during installation of a utilityscale solar project by automating parts of the operation. Robotic machinery drives metal piles into the desert to support the racks on which the modules are mounted. Another robot offoads modules, 8 at a time, from a truck and places them on the racking. By the time we have installed 5,000 modules this way, there are no additional economies of scale for the next 5,000, which is why, from a cost perspective, NREL classifes everything above 2 MW as utility-scale.
we can estimate how much electricity our 250 W solar module will actually generate each hour of the year and we can add these up to get the annual total, say 350 kWh. We can then compare this total with what we would have got if our 250 W module had been in standard test conditions for all 365*24 = 8760 hours of the year: 8760 * 250 / 1000 = 2190 kWh. Our location therefore has a “capacity factor” of 350/2190 = 16%, a number that takes into account the fact that it is not generating at night and it is generating below its full potential on cloudy days and when the sun is low in the sky. Solar has a low capacity factor,7 between 10% and 25% depending on geographic location, with tracking systems achieving higher capacity factors than fxed systems. By comparison, the capacity factor of onshore wind is 25–50% and nuclear and coal, which run almost 24/7, have capacity factors approaching 100%.
If the solar generator has access to the real-time market for electric power, it can sell at the going price. The marginal cost of solar is zero, since there is no fuel cost, and hence it does not make sense to switch off a solar farm just because the spot market price is low. Instead, solar generators that want to play in the market may store electricity until prices are high (e.g. by using a battery). Since the price on the wholesale market varies, there is a risk involved in selling electricity this way. We can mitigate that risk by using a power purchase agreement (PPA)
Suppose we build a solar installation with a capacity of 100,000 watts and operate it at full capacity for ten hours, we would generate 100,000 × 10 watt hours of electricity = 1 MWh for which we would be paid $60 in Dubai in 2014. Using 2014 solar capital costs of $1.97/W, our installation would cost $197,000. With a capacity factor of 23% in Dubai, we need to operate our solar generator for 16 years‡ to recover the capital cost, which looks good since solar modules come with a 25–32 year warranty
Even in 2019, with solar capital costs 51% lower, PPAs in Hawaii were over $60/MWh. In 2020, in Southern California, solar PPAs were around $30/MWh.
one advantage of solar is that it can be built fast (e.g. within a year compared with 5–8 years for nuclear).
Some PPAs include escalation clauses specifying a rate of increase of price over future years, which is important to take infation into account and also because the energy yield of solar modules declines at 0.5–1% per year. The escalation rate may be specifed in the PPA contract or may be linked to infation.
in Western Australia, South Africa and Southwest USA, where air-conditioning constitutes a major proportion of demand during the summer, the base price is increased by, say, 20% during summer afternoons. This is to the advantage of solar, which generates well at those times.
Example of PPA
Bidders should bid a base price, which would be increased by 25% from 1 p.m. to 5 p.m. during June– August. A schedule determines the amount of power to be delivered at different times of day and year. For instance, it could require 10 MW peak power with 30% of the peak being required between 10 p.m. and 7 a.m. To qualify as a bidder in the auction, a company would have to show that they have the right to purchase adequate land and also that planning permission is available. A deposit of $5m is required on signing, which will be refunded in $1m increments for each 2 MW commissioned (i.e. made operational) within 18 months. The PPA price will be escalated at 1% below the electricity price index for the region in future years.
Instead of selling to the grid, some solar projects sell to corporate customers (Amazon, TSMC). Datacenters are a growth market for electricity and examples from Sweden, Quebec and Iceland are also described in another sidebar.
It costs more to transport electricity over transmission lines than to transport data over optical fiber (and the delay is negligible).
Off-site projects can also be located where land is low cost so long as the grid has suffcient capacity to receive the power generated.
They want to be able to say that the solar project would not have been built without the PPA that they signed. This is known as “additionality”.
The capacity factor for a given location allows us to estimate the total electricity that can be generated in a year. This can be converted to a dollar value if the selling price of the electricity is fat rate. If not, we use satellite observations and ground-based equipment that give hourly measurements of the amount of solar radiation throughout the year from which commercial software estimates the electricity generation each hour of the year. This can be mapped on to the time-varying value of that electricity either from a PPA or from the wholesale market price to estimate the annual revenue to be expected from a solar project.
Pg 56 flow diagram
It’s nothing new for organizations to generate their own electricity. By law, hospitals must have back-up power generators in case of a black-out. Somewhere in the basement, they probably have a gas turbine that an engineer checks out every month but which is very rarely used.
Heavy industrial sites such as metal refneries, petrochemical plants and cement factories consume a lot of electricity and many of them do not have a lot of space to install solar modules
Commercial buildings consume less power per square meter and have space to install solar modules on the roof and on awnings over adjacent parking lots.
Demand charges can be quite high (e.g. $10/kWh, compared with $0.05/kWh for electricity), and are designed to provide an incentive for customers to fatten their load profle.
The customer acquisition cost for residential customers is high, distributing fyers and knocking on doors costs $300−$400 per customer who signs up.
A super interesting deep dive into the economics of developing and operating solar modules. I learned a lot about the industry and feel better educated as I continue my exploration!