Wind and solar photovoltaics (PV) are currently the fastest-growing sources of electricity globally. In 2015, their additional annual generation met almost all incremental demand for electricity.
This is the executive summary of a new report from the International Energy Agency
Variable renewable energy (VRE) costs, especially for solar PV and land-based wind, have fallen dramatically in recent years thanks to a combination of sustained technological progress, expansion into newer markets with better resources, and better financing conditions.
As a result, between 2008 and 2015, the average cost of land-based wind decreased by 35% and that of solar PV by almost 80%. A “next generation” phase of deployment is emerging, in which wind and solar PV are technologically mature and economically affordable.
The success of these sources is driving change in power systems around the globe. Electricity generation from both technologies is constrained by the varying availability of wind and sunshine, which makes the output of VRE sources fluctuate over time. The degree to which this poses a challenge depends on the interplay of several factors which vary by country.
For this report, detailed case studies were conducted for Brazil, China, Denmark, Indonesia, Mexico and South Africa.
These countries are at very different stages of VRE uptake (Fig. 1) and also highlight the diversity of contexts for VRE integration. Some systems have only one resource available in abundance (solar in Indonesia, wind in Denmark), while others enjoy high quality resources of both wind and solar (such as South Africa and Mexico). In South Africa, power demand matches closely with wind and solar availability, both across the day and across seasons.
In Denmark, by contrast, demand does not naturally correspond well with VRE, triggering the need for additional flexible resources to be mobilised for system integration. Brazil can rely on the abundant availability of flexible hydroelectric generation to complement wind and solar power well, while the seasonal generation profile of new run-of-river plants harmonise with the country’s wind resources (wind is more prevalent when less water is available).
Mexico, China and Brazil are expected to double their VRE share to reach about 10% of annual generation by 2021; South Africa is forecasted to triple its VRE share, reaching 6%; Denmark grows to reach 60% and Indonesia hardly taps into its VRE potential.
Another highly system-specific property is the evolution of electricity demand and the general need for investment in the power sector. Where demand is growing or a large number of resources are being retired because they have reached their technical lifetime, it is often less challenging to scale up VRE.
For example, South Africa and Indonesia both face a strong requirement for new generation capacity. This means that VRE can be scaled up without putting too much economic pressure on existing generators. By contrast, systems such as those in Denmark and (more recently) China face an environment in which sufficient capacity is in place to meet demand over the coming years. Under these circumstances additional VRE generation displaces existing resources.
At a more general level, the difficulty (or ease) of increasing the share of variable generation in a power system depends on the interaction of two main factors: the properties of VRE generators and the flexibility of the power system into which they are deployed. Flexibility refers to the ability of a power system to maintain reliable operation even in the face of large swings in the supply and demand balance. Power plants, demand-side resources, electricity storage and grid infrastructure can all boost system flexibility.
Integrating the first few percentage points of VRE into annual generation poses few technical and economic challenges, as long as a few basic rules are followed. This finding is in sharp contrast to the initial concerns expressed when wind and solar PV contributed fairly little to power generation.
This means that in systems that currently have very little VRE penetration, such as Indonesia, scaling up is possible with few technical problems in the coming years. This is because traditional power systems already command a significant amount of flexibility to balance power demand. Differently put: the same resources that are used to balance demand may initially be mobilised to integrate VRE.
Where the share of annual electricity generation provided by VRE is growing beyond a few percentage points, power system operation and planning are being upgraded and adapted to accommodate VRE on the grid. Such steps will be a priority for Brazil, China, Mexico, and South Africa in the coming years. As VRE enters its next generation of deployment, the issue of system and market integration becomes a critical priority for renewables policy, and energy policy more broadly.
Once VRE becomes a main source of power generation – as is already the case in Denmark – a comprehensive and systemic approach is the appropriate answer to system integration, best captured by the notion of transformation of the overall power system. This requires strategic action in three areas (Fig. 2):
Achieving power system transformation successfully also requires a shift in the economic assessment of VRE. The traditional focus on the levelised cost of electricity (LCOE) is no longer sufficient. Next-generation approaches need to factor in the system value (SV) of electricity from wind and solar power.
SV is defined as the overall benefit arising from the addition of a wind or solar power generation source to the power system; it is determined by the interplay of positives and negatives. Positive effects can include reduced fuel costs, reduced costs from lower emissions of carbon dioxide (CO2) and other pollutants, reduced need for other generation capacity and possibly grid infrastructure, and reduced losses. On the negative side are increases in some costs, such as higher costs of cycling conventional power plant and for additional grid infrastructure, as well as curtailment of VRE output due to system constraints.
SV provides crucial information above and beyond generation costs; in cases where SV is higher than the generation cost, additional VRE capacity will help to reduce the total cost of the power system. As the share of VRE generation increases, the variability of VRE generation and other adverse effects can lead to a drop in SV.
It is important to distinguish between the short-term and long-term SV of VRE. In the short term, SV is strongly influenced by existing infrastructure and the current needs of the power system. For example, if new generation is needed to meet growing demand or retirements (as in South Africa) SV will tend to be higher. By contrast, the presence of large amounts of relatively inflexible generation capacity (as is the case in Germany) can lead to a more rapid SV decline in the short term. For long-term energy strategies, the long-term SV is most relevant.
This accounts for both fuel savings and capital investment. Power system transformation aims to maximise the SV of VRE even at large shares. A crucial component to achieve this is system-friendly VRE deployment.
System-friendly VRE deployment: maximising the value of wind and solar power
Wind and solar power can facilitate their own integration by means of system-friendly deployment strategies. The case study analysis has revealed examples of good policy practice for mobilising this contribution, as detailed in this report. However, it has also revealed room for improvement.
The fact that VRE is often not seen as a tool for its own system integration has historic reasons. Policy priorities during the early days of VRE deployment were simply not focused on system integration. Instead, past priorities could be summarised as maximising deployment as quickly as possible and reducing the LCOE as rapidly as possible.
However, this approach is not sufficient at higher shares of VRE. Innovative approaches are needed to trigger advanced deployment and unlock the contribution of VRE technology to facilitating its own integration.
A number of general principles apply to the successful design of policies to stimulate system-friendly deployment and maximise SV. These differ for large-scale and distributed VRE plants.
Further detail on the different aspects of system-friendly deployment can be found in the report.
Distributed resources: focus on regulation of local grids and retail prices
A number of policy and market design options can enhance the system-friendliness of distributed resources. Sustaining the safe and reliable operation of the local power network in the face of rising VRE uptake will require up-to-date and technology-specific grid codes for low- and medium-voltage connections. Retail prices should give the right incentives to both network users and distributed energy resources, in a time- and location-specific manner.
In particular, network tariffs need to cover the costs of infrastructure and should send a signal for efficient use of the network, as well as minimise the cost of future investment. This needs to be balanced with other policy objectives, such as economic development in rural communities.
In the context of rising self-consumption, tariff reform is likely to be required. For example, the introduction of demand charges that accurately reflect a customer’s contribution to peak demand in a local distribution grid can be an appropriate way of ensuring fair charges for all users of the network.
In addition, the gradual introduction of time-based pricing to reflect the time-dependent value of power production should be encouraged.
Centralised resources: enhancing remuneration schemes
Reflecting SV in policy frameworks requires striking a delicate balance. On the one hand, policy makers should seek to guide investment towards the technology with the highest SV compared to its generation costs. On the other hand, calculating the precise SV can be challenging and, most importantly, current and future SV will differ.
In practice, exposure to short-term market prices can be an effective way to signal the SV of different technologies to investors. However, the current SV of a technology can be a poor reflection of its long-term value. This is due to transitional effects that can be observed in a number of countries where VRE has reached high shares.
For example, in European electricity markets the combined effect of renewable energy deployment, low CO2 prices, low coal prices and negative/sluggish demand growth (slow economic growth, energy efficiency improvements) are leading to low wholesale market prices. In turn, these low prices mean that any new type of generation will only bring limited cost savings and will thus have a low short-term SV.
Even where electricity demand is growing more rapidly, investments based purely on expected short-term wholesale power prices face multiple challenges. Because wind and solar power are very capital intensive, such challenges will directly drive up the cost of their deployment, possibly widening the gap between SV and generation costs. In addition, current market price signals may be a poor indicator of SV in the longer term.
Consequently, mechanisms are needed to provide sufficient long-term revenue certainty to investors. At the same time, such mechanisms need to be designed in a way that accounts for the difference in SV between generation technologies. A number of strategies have emerged to achieve this.
Two relevant examples are market premium systems, which reward VRE generators which generate higher-than-average value electricity, and advanced auction systems, such as the model recently introduced in Mexico, which selects projects based on their value to the system rather than simply on generation costs.
As next-generation wind and solar power grow in the energy mix, a focus on their generation costs alone falls short of what is needed. Policy and market frameworks must seek to maximise the net benefit of wind and solar power to the overall power system.
A more expensive project may be preferable if it provides a higher value to the system. This calls for a shift in policy focus from generation costs to SV.
Next-generation wind and solar power calls for next-generation policies
Action across five strategic areas is needed:
Power system transformation
Advanced VRE technology
This executive summary is published here with permission.
Contact Marc-Antoine Eyl-Mazzega, IEA, Marc-Antoine.firstname.lastname@example.org