Achieving high levels of renewable integration is technically feasible—however, economic impacts vary dramatically depending on policies, regulations, investments, and many other issues that present hard choices. Objective analyses of the trade-offs often rely on market studies using production cost modeling (PCM) to perform cost/benefit analyses of future scenarios and to compare likely impacts of alternative choices. PCM attempts to represent the actual physics and business practices that govern system operations through power-system fundamental analysis. This is in contrast to technical analysis that attempts to forecast future conditions through trend analysis and through relationships with explanatory variables that may not be based on a “fundamental” description of physics or practice.
Cost/benefit results of PCM are only as good as the models’ representation of real-world power systems and the assumptions used to build those models. Today’s traditional simplifying assumptions are no longer valid, and new planning methods have not adapted yet to the impacts of renewable integration, particularly with regard to the role of forecast uncertainty, real-time variability, and resource diversity. Power-systems operations and reliability planning have always been driven by the impacts of uncertainty, but the nature of uncertainty posed by renewables is different. From an operational perspective, traditional sources of uncertainty can be classified as either high MW impact with low probability (e.g., forced outages) or low MW impact with high probability (e.g., load-forecast error).
Renewable generation presents planners and operators with uncertainty that can be both high impact and high probability. This is amplified by the weather-dependent correlation with load, forced outages, and other traditional sources of uncertainty. Better fundamental models are needed to anticipate how this uncertainty will impact future power-system operations. This is especially important because past operating experience managing power-systems is not sufficient to anticipate operations with high levels of renewable generation and associated changes to system controls and other resource capabilities.
Getting It Right: Economic Planning
Established methods used to manage power-system uncertainty have relied on a clearly defined set of reserve requirements to address traditional sources of uncertainty: contingency reserves for low-probability high-MW events, and regulating reserves for high-probability low-MW events. With renewable integration, new types of reserves are being explored to address the uncertainty of renewables: load-following reserves, ramping reserves, and other types of ancillary services. Though we have increasing experience with managing renewables, we have yet to settle on best practices for how to manage uncertainty in future grids with high renewables. This is illustrated by efforts to use regulating reserves to manage uncertainty of renewables, but the high costs of regulating reserves have pushed operators to identify alternatives.
Without better models of uncertainty and system operations, we cannot compare the impacts of changing reserve requirements, nor can we objectively evaluate the costs and benefits of renewable generation, transmission, ancillary services, markets, controls, regulations, or public policies. New models are essential to make effective choices about how we design and operate future power systems. If we fail to appropriately incorporate the impacts of uncertainty and operating practices in our economic planning studies, we will make poor investment and market-design choices that make renewable integration excessively costly or unaffordable, potentially even blocking our ability to achieve high levels of integration.
An Example: Assessing the Value of Transmission
A recent report, The Value of Diversifying Uncertain Renewable Generation through the Transmission System, illustrates the impact of day-ahead forecast uncertainty of renewable generation and load on assessing the value of transmission facilities. In summarizing their results, the authors identify that
When real-time uncertainties of renewable generation are taken into consideration, the benefit of geographic diversification through the transmission grid are 2 to 20 times higher than benefits quantified only based on “perfect forecasts” under day-ahead market conditions.
The following figures from the report show the impact of uncertainty on simulated curtailments of renewables and on the economic benefits (i.e., system-wide operating-cost savings) of transmission facilities interconnecting two regions with diverse renewable generation patterns. For annual curtailment, solid bars of both colors represent curtailment with day-ahead “perfect foresight”; hashed bars of both colors represent the additional curtailment captured by also simulating uncertainty between day-ahead and real-time.
These show that neglecting impacts of uncertainty in PCM causes results that significantly underestimate the curtailment of renewables and other system operating costs. This then leads to significant underestimation of the benefits from interconnecting systems. Unfortunately, as the authors note, “little work has been done to quantify (1) how the need to integrate growing amounts of renewable generation increases the benefit of expanding regional and interregional transmission capabilities; and (2) how the day-to-day renewable generation forecasting uncertainties further add to these benefits.”
While this analysis was focused on the system-wide benefits of transmission, the results speak broadly to the necessity of incorporating uncertainty when assessing the value of other services that provide flexibility or reduce uncertainty, including: the value of better forecasts, new and re-rated facilities, modified operating policies, load participation, and many other choices available to adapt power grids to high renewables.
Getting to High Renewables Requires Significant Changes to Planning Methods
A cost-effective and reliable transition to a high-renewables future requires significant changes to our planning methods. Planning must incorporate new trade-offs driven by the economic impacts of forecasting uncertainty and real-time variability. If we do not reflect these impacts in economic planning studies, we will not achieve high levels of renewable integration, or we will achieve this only with high costs and degraded system reliability. Models that can more-accurately simulate impacts of variability and uncertainty will allow us to achieve high levels of renewables affordably and reliably.
Polaris Systems Optimization