(667a) Grid-Level “Battery” Operation of Chemical Processes with Engagement in Short-Term Electricity Markets

The U.S. electric power generation sector is experiencing a transition from a set of concentrated resources to a more distributed system. This is in part driven by the deregulation of the electricity markets and the influx of renewables into the energy mix [1, 2]. Accompanying this transformation is the need for a robust grid, capable of handling variability on both the supply and demand sides. Mechanisms such as energy storage and demand response (DR) are being discussed and implemented by diverse stakeholders as a means to bridge instantaneous demand and supply mismatch and thus enhance grid operations.

Amongst such stakeholders are industrial consumers, who could potentially contribute significantly to peak load reduction by engaging in DR [3]. Industrial engagement in the operation of the power grid can occur over multiple time scales. Involvement in long-term (day-ahead) energy markets (LTM) can lead to considerable minimization of energy costs for the consumer subject to time-varying power tariffs [4]. To this end, industrial consumers typically reduce power use during peak-tariff hours by diminishing their production rates. Production is increased at off-peak hours, and the excess product is stored to meet demand during the subsequent peak period.

Participation in short-term markets (STMs) can be even more lucrative [5], but comes with some significant challenges. STMs are highly volatile, with conditions varying over time scales in the order of minutes (or even seconds). Since the dynamics of process systems are inherently slower, with time constants in the order of hours, altering production rates to capture fast changes in market conditions will imply a highly transient plant operation, with potentially abrupt changes in the production schedule (including e.g. schedules optimized to take advantage of the LTM conditions). As a consequence, it is imperative that, i) the production schedule be re-optimized to account for changes in STM conditions, and, further, ii) the process dynamics must be accounted for in the scheduling calculation [4] to guarantee that production schedules are feasible.

In our previous work [4], we addressed the issue of engaging chemical processes in LTM DR by proposing a novel scheduling framework that incorporates a representation of the process dynamics. The framework is predicated on the use of low-order representations of the scheduling-relevant dynamics of a process, which we refer to as “scale-bridging models” (SBMs) [6]. In this work, we extend this framework to account for participation in STMs. To this end, we formulate a two-tier optimization problem. In the top level problem, a base load profile is computed, that maximizes process profit for a fixed time horizon based on day ahead market (DAM) prices. The lower tier problem is solved over a shrinking time horizon, aimed at determining in every short-term time block any additionally available DR capacity for the process (which we define as the “DR headroom”) and its cost relative to the base cost. Subsequently, process operators can offer the DR headroom for sale or auction in the open market. Thus, the process is acting as a grid-level “battery” in the STM, offering load reductions and earning revenue. Our approach accounts for any changes in the remaining long-term schedule, including potential future increases in production to compensate for the instantaneous decrease dictated by STM conditions. We illustrate our developments with a case study, showing that the potential for increased earning for chemical processes is significant.

 

References

1. Lacy, V., Matley, R., & Newcomb, J. (2012). Net Energy Metering, Zero Net Energy and the Distributed Energy Resource Future: Adapting Electric Utility Business Models for the 21st Century.
2. Davoli, F., Repett, M., Tornelli, C., Proserpio, G., & Cucchietti, F. (2012). Boosting energy efficiency through smart grids. International Telecommunication Union (ITU).
3. 2016 Assessment of Demand Response and Advanced Metering Staff Report. Federal Energy Regulatory Commission.
4. Pattison, R. C., Touretzky, C. R., Johansson, T., Harjunkoski, I., & Baldea, M. (2016). Optimal Process Operations in Fast-Changing Electricity Markets: Framework for Scheduling with Low-Order Dynamic Models and an Air Separation Application. Industrial & Engineering Chemistry Research, 55(16), 4562-4584.
5. Dowling, A. W., Kumar, R., & Zavala, V. M. (2017). A multi-scale optimization framework for electricity market participation. Applied Energy, 190, 147-164.
6. Pattison, R., Touretzky, C. R., Johansson, T., Baldea, M., & Harjunkoski, I. (2016). Moving horizon scheduling of an air separation unit under fast-changing energy prices. IFAC-PapersOnLine, 49(7), 681-686.