A road map for large scale renewables integration1 July 2014
The cost effective integration of variable renewable energy (VRE) has become a pressing challenge for the energy sector. A new study by the IEA suggests that the necessary integration is possible, but may need a system-wide transformation. It describes measures by which each country could deal with its own particular circumstances.
Wind power and solar photovoltaics (variable renewable energy, or VRE) are likely to be crucial in meeting future energy needs while decarbonising the power sector. Deployment of both technologies has expanded rapidly in recent years, one of the few bright spots in an otherwise bleak picture of clean energy progress, and IEA projections indicate that this trend will continue for decades. However, the inherent variability of wind power and solar PV raises unique and pressing questions. Can power systems remain reliable and cost-effective while supporting high shares of variable renewable energy (VRE)? And if so, how?
Working from a thorough review of the integration challenge, the IEA has published a guide, The Power of Transformation*, that gauges the economic significance of VRE integration impacts, highlights the need for a system-wide approach to integrating high shares of VRE, and recommends how to achieve a cost-effective transformation of the power system. It confirms that integrating high shares - that is, 30 % of annual electricity production or more - of wind and solar PV in the power system of any country can come at little additional cost in the long term if the route taken is carefully chosen.
This guide summarises the results of the third phase of the Grid Integration of VRE (GIVAR) project, undertaken by the IEA over the past two years. It is rooted in a set of seven case studies from 15 countries on four continents. It deepens the technical analysis of previous IEA work and lays out an analytical framework for understanding the economics of VRE integration impacts. Based on detailed modelling, the impact of high shares of VRE on total system costs is analysed. In addition, the four flexible resources which are available to facilitate VRE integration - generation, grid infrastructure, storage and demand side integration - are assessed in terms of their technical performance and cost-effectiveness.
Transforming power systems
Costs depend on how flexible the system currently is and what strategy is adopted to develop flexibility in the long term. Managing this transition will be more difficult for some countries or power systems than others, the study finds. 'Integrating high shares of variable renewables is really about transforming our power systems' said IEA executive director Maria van der Hoeven at the launch of the report. 'This new IEA analysis calls for a change of perspective,' she said. 'In the classical approach, variable renewables are added to an existing system without considering all available options for adapting it as a whole. This approach misses the point. Integration is not simply about adding wind and solar on top of 'business as usual'.
Currently, wind and solar PV account for about 3% of the world generation total, but some countries the figure is far higher. In Italy, Germany, Ireland, Spain, Portugal, and Denmark, they amounted to between 10 and 30% or more in 2012 (Figure 1).
The report finds that, for any country, integrating the first 5-10 % of VRE generation poses no technical or economic challenges, provided that three conditions are met: uncontrolled local 'hot spots' of VRE deployment must be avoided, VRE must contribute to stabilising the grid when needed, and VRE forecasts must be used effectively. These lower levels of integration are possible within existing systems because the same flexible resources that power systems already use to cope with variability of demand - flexible power plants, grid infrastructure, storage and demand-side response - can also help integrate the variability of wind and solar.
But going beyond the first few percent to reach shares of more than 30 % will require a transformation that has three main requirements: deploying variable renewables in a system-friendly way using state-of-the art technology, improving the day-to-day operation of power systems and markets, and finally investing in additional flexible resources.
The challenges of such a transformation depend on whether a power system is 'stable', meaning no significant investments are needed in the short term, or 'dynamic' which requires significant investments short term, to meet growing power demand or replace old assets.
Varying responses globally
In stable systems, such as those in Europe, the existing asset base will help to provide sufficient flexibility to increase VRE generation further.
However, in the absence of demand growth, increasing VRE generation in stable systems inevitably works to the detriment of incumbent generators and puts the system as a whole under economic stress.
This outcome is based on fundamental economics; market effects are not a consequence of variability alone. The transformation challenge in stable systems is twofold: scaling up the new, flexible system while scaling down the inflexible part of the old. Governments with stable systems face tough policy questions about how to handle the distributional effects, in particular if other power plants need to be retired before the end of their lifetimes and, if so, who will pay for stranded assets. Meeting these challenges will only be possible through a collaborative effort by policy makers and the industry. In any case, "these surmountable challenges should not let us lose sight of the benefits renewables can bring for energy security and fighting dangerous climate change. If OECD countries want to maintain their position as front runners in this industry, they will need to tackle these questions head-on," commented Ms van der Hoeven. By contrast, in 'dynamic' power systems such as in India, China, Brazil and other emerging economies, wind power and solar PV can be cost-effective solutions to meet incremental demand.
VRE grid integration can - and must - be a priority from the onset. With proper investments, a flexible system can be built from the very start, in parallel with the deployment of variable renewables. Ms van der Hoeven - "Emerging economies really have an opportunity here. They can leap-frog to a 21st-century power system".
The additional cost of reaching high shares of VRE critically depends on how well the system is adapted as a whole.
This finding is the result of detailed economic modelling at hourly time resolution for a test power system. When allowing for full transformation of the entire system ('transformed case'), a share of 45% VRE in annual electricity generation (15% solar PV, 30% onshore wind) increases total system costs of the test by USD 11/MWh, compared to a share of 0% VRE (Figure 2). This corresponds to an increase of about 10%-15%. At 30% VRE system costs increase by USD 6/MWh, or about 7%.
The transformation includes a re-optimised mix of dispatchable power plants as well as additional flexibility in the form of demand side response using thermal energy storage. It also assumes an optimised strategy for managing grid infrastructure and uses current wind and solar PV technology costs, and a carbon price of $30/tonne.
In the extreme case that a share of 45% is added overnight without changing the power plant mix and optimising grid infrastructure ('legacy case'), total system costs increase by as much as USD 33/MWh, or about 40%.
The revised IEA Flexibility Assessment Tool (FAST2) assesses the technical ability of power systems to absorb higher shares of VRE, given today's levels of system flexibility in different case-study regions. Penetration levels of 25% (inflexible systems, eg Japan) to 40% (flexible systems, eg Brazil) of annual generation are technically feasible, if there is sufficient grid capacity.
These shares can be increased further to >50% if a small amount of unused VRE generation is accepted (curtailment). However, mobilising system flexibility to its technical maximum can be considerably more expensive than least-cost system operation.
Medium-term estimates (to 2018) expect VRE shares in electricity generation in the highest level VRE countries to be approximately: Denmark (45%), Ireland (30%), Portugal (25%), Germany (20%), Spain (20%) and Italy (15%). India, Japan and Brazil are projected to have annual generation shares of around 5% in 2018, up from less than 2% in 2012. Looking ahead to 2035, the IEA World Energy Outlook scenario featuring the most aggressive deployment of wind power and solar PV sees VRE shares in power generation at 31% in OECD Europe, 20% in the USA, 19% in Japan, 16% in India and 7% in Brazil.
Transformation: Four resources
Countries with stable power systems should seek to maximise the contribution from existing flexible assets and consider accelerating system transformation by decommissioning or mothballing inflexible capacities that are surplus to the system. But those with dynamic systems should approach system transformation as a question of holistic, long-term system development from the outset. This requires the use of planning tools and strategies that appropriately represent the potential of VRE for a cost-effective, low-carbon outcome.
There are four fundamental flexible resources that can facilitate VRE integration by increasing the flexibility of the power system: flexible power plants, grid infrastructure, storage and demand side integration. A simplified metric, the levelised cost of flexibility (LCOF) assesses the cost of the different flexible resources, as follows:
- Flexible generation. Reservoir hydro and certain gas power plants are extremely flexible technically and are among the cheapest options available. The additional cost to obtain flexibility from these sources ranges from USD 1/MWh to USD 5/MWh. LCOF can be much higher if inflexible technologies are used to ramp generation up or down quickly and frequently or if flexibility provision leads to a reduction in the capacity factor, in particular for power plants that were designed to operate around the clock.
- Grid infrastructure. The only flexibility option which brings a significant double benefit: it is a precondition for connecting more distant resources, and it helps smooth the variability of wind and solar PV by aggregating their output over larger areas. Transmission can be a relatively low-cost option, as low as USD 2/MWh; its LCOF shows a high sensitivity to utilisation rate.
- Distributed power. The additional costs for accommodating small-scale solar PV are moderate - as low as $1/MWh for a PV system size featuring 2.5 kW per household - if the grid is planned properly from the onset. Costs for retrofits may be considerably higher.
- Electricity storage can provide a broad range of different services, but it remains roughly ten times more expensive than other options. In exceptionally favourable cases, pumped hydro plants can deliver flexibility for $20/MWh, if existing hydro can be used. Costs for new plants range from $30/MWh to more than 200/MWh. Battery technologies are more expensive than pumped hydro storage. For example, using lithium-ion batteries may cost $200/MWh to more than USD 800/MWh.
- Demand-side integration can provide flexibility cost-effectively. Including the cost for smart meters, distributed thermal storage can deliver flexibility for as little as approximately $7/MWh. Once communication and control infrastructure is in place and demand-response capabilities have been streamlined into design of appliances, additional costs can be negligible.
The cost of flexibility is only half the picture. The value of the flexibility that each source can provide is equally important. If a high-cost resource provides a very high-value type of flexibility, it can still be cost-effective. Using two different economic simulation tools, the IEA report investigated the cost benefit of different flexible resources.
Demand-side integration (in particular distributed thermal storage) shows superior cost-benefit performance compared with other flexibility options. However, a degree of uncertainty exists over its full potential in real applications. Cost-benefit profiles of storage are less favourable, reflecting higher costs. However, where multiple benefits align (eg avoiding grid investments, providing reserve capacity), storage can be cost-effective today. Potential cost reductions for storage merit further investigation.
Interconnection allows a more efficient use of distributed flexibility options and generates synergies with storage and demand-side integration (DSI). Modelling for the North West Europe case study showed a favourable cost-benefit of significantly increased interconnection. However cost-benefit analysis of retrofitting existing power plants to increase flexibility was less certain, showing a wide range of outcomes, driven by project-specific costs.
The Power of Transformation - Wind, Sun and the Economics of Flexible Power Systems is on sale from the IEA bookshop