substitution with electricity versus substitution within low carbon resources
Electricity is a high quality form of energy. It can carry vast amounts of energy across continents or allow the fine grained control of information. The high quality nature of electricity makes it highly substitutable. Electricity allows decarbonization of transport and heating. Low carbon electricity can power cars or to heat our homes.
Unfortunately, low carbon electricity resources are themselves not easily substitutable. Different low carbon resources offer fundamentally different value to the grid. To use one resource in place of another is possible, but ends up being more expensive the more substitution you do.
Jenkins groups low carbon resources into three classes
- fuel saving variable - wind & solar
- fast-burst - energy storage & demand response
- flexible base - nuclear, carbon capture & storage, electricity to gas & seasonal storage
Wind and solar displace burning fossil fuels. Fast-burst resources displace capacity, avoiding the need to build, operate and maintain standby capacity. Flexible base resources offer both the ability to offset fossil fuels and to reduce capacity.
Jenkins reports that progress in variable renewables is roughly on track, but progress in flexible base resources is behind. Mature technologies such as nuclear or CCS are largely stagnant. Newer technologies such as underground storage or electricity to gas are unproven on the scales required to make a real difference.
The troubles of nuclear in the West are well documented. Today large scale nuclear plants are only being built in planned economies (i.e. China & the Middle East). Jenkins notes that investors struggle to handle the large absolute values of capital required to build nuclear in market based economies, even if the
$/GW of a project is attractive.
Today’s low interest rates should make investing in nuclear more attractive, yet nuclear in the West is reducing in capacity. The West has largely forgotten how to build nuclear on time and on budget, with South Korea becoming the world leader in nuclear builds.
Smaller nuclear plants can help make the investment easier to swallow. Smaller plants also reduce construction and financing risks, and allow manufacturing to be done in factories (rather than be assembled on site).
variable renewables peak
Jenkins shows the optimal relative mixture of low carbon resources as a function of carbon limits. Figure 1 below shows that initially it is optimal to focus on increasing the relative share of variable renewables on the grid.
It also shows that later on the grid requires other types of resources to deliver the lowest cost electricity supply. Eventually the costs of operating a variable renewable grid become so large that more expensive but dispatchable low carbon generation (i.e. nuclear) becomes the optimal economic decision.
This means that we must continue to develop, support and improve fast-burst and flexible low carbon resources alongside variable renewables. Not because we need them today but because we need them tomorrow.
value versus cost
Jenkins explains the historical mental model for supporting clean technology. Subsidies allow a technology to reduce its costs and move up it’s experience curve. The path often includes developing economies of scale (i.e. in manufacturing or supply chains) and accumulation of iterative ‘learning by doing’ improvements. Once the cost of the technology is low enough then the technology can stand on it’s own two feet without subsidy.
The new mental model is that as renewable penetrations increase we see both decreases in cost and a concurrent decrease in value. Because renewables generate at the same time, the oversupply during these times drives down electricity prices. This driving down of prices is a signal that the grid isn’t valuing generation during these time periods. If the value of electricity generated by renewables is very low then even very cheap plants won’t get built.
There are four reasons why the value of wind & solar decline as renewables penetration increases
- offsetting less fossil fuels
- offsetting less capacity requirements (i.e. avoided costs of building dispatchable generation)
- increased curtailment
- increased grid integration costs - reserve requirements & network costs
diversification offers the lowest cost decarbonization
In the context of the type of optimization models Jenkins develops this makes complete sense. Introducing additional constraints onto a linear program will only ever end up with the same or a worse result.
Not allowing nuclear limits us to an equal or worse off situation than allowing nuclear. If a nuclear free pathway was optimal then both models will find the same solution. The fact is that they don’t. Jenkins finds a strong consensus in literature that a diversified mix of low carbon resources offers cheapest deep decarbonization. Dispatchable low carbon resources significantly reduce the cost and technical resources of deep decarbonization.
Not using nuclear would require a massive (roughly double) build out of variable renewables capacity. Not only do we need more capacity but the global capacity factor
MWh / MWh maximum will be low. The capacity will also have a very high cost per utilised output
$/MWh. The low energy density nature of variable renewables also means greater land use impacts for a variable renewables scenario only solution.
A variable renewables only solution also requires long duration energy storage. In a review of literature, Jenkins finds a seasonal storage requirement for 8-16 weeks worth of US electricity consumption. To put this in context - the ten largest pumped hydro plants in the US currently have around 43 minutes of storage.
Thanks for reading!