Nuclear and Renewables: a deep decarbonization pathway

Leonam dos Santos Guimaraes

Eletrobras Eletronuclear S.A.

The 2015 jointly report on Projected Costs of Generating Electricity [1], International Energy Agency (IEA) and Organization for Economic Cooperation and Development Nuclear Energy Agency (OECD/NEA) provided evidence on two key points:

?      Despite recent high-cost projects in Western countries, most new nuclear plants have a Levelized Cost of Electricity (LCOE) comparable to any other generation source, including most Variable Renewable Energy (VRE). LCOE meets all costs, including Capex and Opex, until a new plant is connected to the grid; and

?      LCOE for VRE did not take into account the system costs that consumers would be required to pay, such as network upgrades to accommodate a distant generation from consumer centers, low VRE predictability balancing and frequency control and backup and/or storage of electricity to compensate for this variability.

System costs are often divided into the following four broadly defined categories of profile costs (also referred to as utilization costs or backup costs), balancing costs, grid costs and connection costs:

?      Profile costs refer to the increase in the generation cost of the overall electricity system in response to the variability of VRE output.

?      Balancing costs refer to the increasing requirements for ensuring the system stability due to the uncertainty in the power generation (unforeseen plant outages or forecasting errors of generation).

?      Grid costs reflect the increase in the costs for transmission and distribution due to the distributed nature and locational constraint of VRE generation plants.

?      Connection costs consist of the costs of connecting a power plant to the nearest connecting point of the transmission grid.

Using LCOE to compare generation costs has become widespread practice. However, the approach based on comparisons of LCOE associated with different generation technologies, or any other measure of total life cycle production costs per MWh provided, does not take in account different system costs, effectively treating all generated MWh, regardless of source, as a homogeneous product, i.e. a commodity, governed by a single price.

The criticism is technical and the fundamental objection is that cost does not measure value. Power generation occurs at different times and in different places, having different values at each moment and in each place. It would be like saying that a car costs a lot more than a bicycle, so we should all buy bicycles. Nevertheless, this disregards that car and bicycle are providing services of different natures.

LCOE analysis does not include environmental and social externalities such as waste disposal, pollution reductions and material resources land use. Excluding marginal externalities, LCOE contradicts a central point for the consideration of clean energy technologies, which is the very impact of these externalities [3].

OECD/NEA presented on January 2019 the conclusions of its latest publication on the costs of decarbonized systems with high shares of nuclear and renewables [2]. Aimed at policy makers, the new study raises its main questions by assessing the choices that would effectively achieve deep decarbonization, that is, an emission of less than 50gCO2/kWh of a future electrical system. Are they:

?      What is the most economical combination to achieve a decarbonization target with a certain VRE share?

?      What technologies are available, what are their reasonably expected costs, and what are their effects on the overall reliability of the electrical system?

?      What policies will lead to the long-term investments that deep decarbonization requires?

The study compares different systems that reach the same 50gCO2/kWh, which is roughly the goal of the Paris Agreement to achieve, by 2050, the climate change scenario of 2 degrees Celsius. It considers a “greenfield” approach, i.e., defining from scratch an entire system that would be capable of generating 540 GWh in a year to avoid many assumptions from existing technologies or systems.

The baseline scenario comprises some hydroelectric, nuclear and some gas plants. It is the least expensive in LCOE, network and system costs; has no VRE (its LCOE remains higher than nuclear to achieve a certain generation in MWh) or dedicated storage for VRE; and has a robust network, setting a benchmark for system costs. It also assumes a carbon price constraint of $35 per ton of CO2, which virtually eliminates any coal-fired thermoelectric generation. This scenario bears some resemblance to the French system, which still provides the cheapest electricity to end consumers in Europe, even with all subsidies and taxes included.

Four scenarios explore a growing share of VRE in the generation mix - 10%, 30%, 50% and 75% respectively - and analyze the generation capacity required along with their LCOE plus any additional system costs compared to the base scenario. In addition, the new study emphasizes the role of “profiling” or “utilization” cost, which reflects the expense of providing complementary generation when VRE participation increases. The more VRE there is in the system, the more expensive this generation is, as system costs increases significantly.

Not surprisingly, by increasing the share of VRE in the electricity generated, the required capacity will grow to more than three times than expected in the baseline scenario. Nuclear power decreases significantly, and gas capacity more than doubles. This decrease is due to the high load variation requirements that the VREs impose on the system. Nuclear can accommodate this, but above a certain level, the load factor is affected in such a way that this source becomes uneconomical. Battery storage plays a very limited role in any of the study scenarios. Its cost is affordable for frequency control or short-term balancing, but it′s hard to find a case this storage could play a role in the medium to long term because due to cost. Therefore, the balance of electricity, where there are no VRE and nuclear is not enough, must come from gas.

The cost of electricity, LCOE plus system costs, increases from US $ 65/MWh in the baseline scenario to US $ 130/MWh, when 75% is VRE. In addition to LCOE, system costs increase almost exponentially with VRE participation, rising from a minimum of US $ 8/MWh in the 10% VRE scenario to US $ 50/MWh in the 75% scenario. Network, balancing, and connection are responsible for one third, and profile ones for two thirds. Most profiling costs come from system de-optimization arising from VRE variability due to the correlation in renewable production hours that is seen in large existing systems. The more VRE is deployed, the more expensive the overhead will be. With a share of these sources greater than 50%, excess capacity, when all types of variable renewable energy are generated,

The study notes the high volatility of electricity prices in the market. Existing markets are mainly based on merit order of marginal cost of electricity generation, but VREs have zero marginal cost. This means that when the share of these sources increases to the level of demand capacity, variable renewable energy may occasionally meet all consumption needs and lead to a market price of $ 0/MWh, along with restriction of some VREs. From 10% to 75% of VRE, the number of hours at a price of $ 0/MWh grows from a few to almost 4,000 a year. In compensation, the number of hours at prices above $ 100/MWh increases. This leads to high volatility and unpredictability of electricity prices.

The NEA/OECD report makes five main recommendations that will be commented in the following:

?      fairly recognize and allocate system costs to the technologies that cause them and promote short-term competitive markets for the economic dispatch of available technologies;

?      encourage new investments in low carbon technologies, providing stability to investors;

?      ensure adequate levels of capacity and flexibility, as well as transmission and distribution infrastructure;

?      implement carbon pricing as the most efficient approach to decarbonizing the electricity supply;

?      create appropriate policies for the rapid deployment of all available low-cost technologies as cost-effectively as possible to decarbonize the electricity sector successfully.

Setting a price for carbon seems obvious. The study shows that $ 35 per ton of emitted CO2 is considered sufficient to eradicate it from all its scenarios. This is not so far from the $ 20 already considered by some countries. The sooner this is achieved, the better, since everyone agrees that there is an urgent need to decarbonize the energy system.

Ideally, policies should be developed to ensure that system costs are well analyzed and allocated to the source that generates them. In the UK, the concept of Equivalent Firm Power [4] was proposed, according to which any Variable Renewable Energy (VRE) source should guarantee its production with some storage for which it would be responsible. In any system, this would be very difficult to implement.

The adequacy of most existing electricity markets may be questioned: the order of merit could have been justified in the past when all sources had comparable Levelized Electricity Costs (LCOE) and were fully exposed to the market. Electricity markets produce situations where prices are zero and there are no longer economic signals consistent with an increasing share of VRE.

In a market where any form of electricity generation is dealt with on its own merits, without any subsidies or priority rights, there will be a need for very clear new regulations. With a high share of VRE, existing markets will be very volatile and will pose high risks to any long-term investment and financing. How can policies be designed to attract investment in this situation?

There is clear evidence that in addition to hydroelectric power with large reservoirs, nuclear is the only low-carbon dispatchable technology, and it is essential, along with variable renewable energy, to obtain a decarbonized electrical system. The cost-benefit ratio for the consumer leads to a balanced system where the value of nuclear energy and the VREs themselves is not destroyed by excessive participation by the latter. Rather than developing public policies that set targets for VRE participation, which will require network capacity, flexibility and infrastructure, it would not be preferable to set carbon generation targets first and then identify which electrical system would provide the best cost-benefit?

When considering the facts about the types of technology; their costs, including system costs; public acceptance; and by assessing the potential for higher electricity prices, policy makers could create the market conditions and rules to find an appropriate path.

There are issues that are beyond the scope of the OECD-NEA study, but they are important for decision makers. One is that in order to accommodate a high share of VRE, the system must develop not only transmission and distribution networks but also incorporate new technologies that do not yet exist to accommodate the fluctuations that VRE generation entails. These costs may have been taken into account in the study, but what about the risks associated with these future technologies? And the reliability of such a system and its resilience?

The study did not examine the material resources required in any scenario. However, this is a topic to consider. In essence, VRE has, in most areas, a limited load factor: to achieve the same generation in GWh, VRE needs three times more capacity than any dispatchable source and would require a lot of storage capacity with, again, a limited load factor. Low energy density VRE implies more building materials (cement, concrete, steel, for example) and more land use for a given lifecycle energy generation. Is this the most efficient way to use the resources the planet can offer?

Another issue to consider is the acceptability of a given scenario. While existing nuclear power generation is generally well accepted, new nuclear power can be a challenge. What about a comparatively large VRE deployment and its impact? What about the acceptability and feasibility of distribution/connection requirements?

Regarding the different scenarios analyzed by OECD/NEA study, a cost-effective low carbon system would probably consist of a sizeable share of VRE, an at least equally sizeable share of dispatchable zero carbon technologies, such as nuclear energy and hydroelectricity. A complementary amount of gas-fired capacity would provide additional flexibility, alongside storage, demand side management and the expansion of interconnections.

Nuclear power will play a key role in future decarbonized systems. Although it reliably produces large quantities of low-carbon, dispatchable energy, it faces issues of public acceptance in many countries. However, this study shows how nuclear power remains an economically viable option to meet severe carbon constraints, despite the economic challenges for some new reactor projects.

The cost advantage of nuclear power is not in its plant-level costs, although they are quite competitive. It does lie in its general benefits to the electrical system. VRE's plant-level costs have fallen dramatically, but its overall system costs are not accounted for as production is aggregated over a limited number of hours. All of these factors must come into play in the decisions of each country. 

Electricity markets are evolving and nuclear energy is following this evolution to meet future requirements: Small Modular Reactors (SMR) development is a promising response. Nuclear energy is well placed to take on these challenges in a collaborative mode, working together with all other forms of low carbon generation, in particular VRE, to achieve the ambitious decarbonization targets countries have set for themselves.

References

[1] INTERNATIONAL ENERGY AGENCY AND NUCLEAR ENERGY AGENCY, Projected Costs of Generating Electricity. (2015)

[2] ORGANIZATION FOR ECONOMIC COOPERATION AND DEVELOPMENT NUCLEAR ENERGY AGENCY, The Costs of Decarbonization: System Costs with High Shares of Nuclear and Renewables. (2019).

[3] GUIMARAES, L., The Levelized Cost of Electricity and its Impact on Energy Transition, CEIRI NEWS (available in Portuguese on https://ceiri.news/o-custo-nivelado-da-eletricidade-e-seu-impacto-na-transicao-energetica/). (2019)

[4] HELM, D., Cost of Energy Review, BRITISH INSTITUTE OF ENERGY ECONOMICS. (2017)