Executive SummaryPower systems of the future are likely to require additional flexibility due to the operating characteristics of many clean energy technologies, particularly those relying on renewable energy sources. This subject has been well studied from an operational perspective, but it has been more difficult to incorporate into capacity expansion models (CEMs) that study investment decisions on the decadal scale. There are two primary reasons for this. First, the necessary input data, including cost and resource projections, for flexibility options like demand response and storage are not widely available and are highly uncertain. Second, it is computationally difficult to simultaneously represent both investment and operational decisions in detail, with the latter being necessary to adequately value system flexibility in realistic systems.The primary purpose of this report is to present new capabilities that were developed for a particular CEM, NREL's Resource Planning Model (RPM), to better reflect the impact of variable wind and solar generation on system operations and resource adequacy, and, complementarily, to model energy-constrained flexibility resources. The additional variable generation modeling capabilities enable a more-accurate representation of the need for additional flexibility in systems with high penetrations of variable generation and, as a result, better reflect the value of technologies that can provide flexibility. We demonstrate the use of the new capabilities by examining the role of two broad technological sources of flexibility in power systems-utility-scale storage and interruptible load-and we model how they might contribute to a range of grid services for operations and planning. A wide range of different energy storage technologies is modeled, including multiple types of battery technologies. We also model interruptible load, a type of demand response that can be used to reduce demand, usually during peak hours, within pre-determined capacity and duration limits.The new capabilities use load duration curves from hourly data spanning a full year, thus enabling RPM to capture the impact of variable generation and the potential value of storage and interruptible load during infrequent, but important periods and hours within the year. The techniques enable RPM to accurately calculate the need for grid services and simultaneously consider, in its investment decision, the potential for storage and interruptible load to provide value to the system in multiple distinct ways: firm capacity for resource adequacy, operating reserve provision, energy arbitrage, and reductions in renewable curtailment. The ability of storage and interruptible load to lower system costs by providing these services, as well as the capital and implementation costs and other characteristics of the flexible technologies, are factored into RPM's economic investment decisions. These new techniques adapted for RPM can also be implemented in other CEMs.Using these capabilities, we analyze deployment of storage and...