DER Forecasting for Electric Distribution System Planning
- 1. Spatial, Dynamic Modeling of the Economics, Adoption and Demand Impact of Distributed Energy Resources to Support Utility Distribution System Planning PRESENTED TO: MARCH 5, 2019
- 2. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Introductions Cory Welch Founder & President James Milford Director of Technology Lumidyne staff have been developing sophisticated energy-focused planning models for electric & gas utilities throughout North America for over a decade. Mr. Welch & Mr. Milford are both former NREL employees who worked on development of NREL’s Stochastic Energy Deployment Systems (SEDS) model, and NREL’s Hydrogen Dynamic Infrastructure and Vehicle Evolution (HyDIVE) model. Expertise in energy efficiency, solar PV, storage, demand response, & electric vehicle modeling • Skills include System Dynamics, technology diffusion, linear & nonlinear optimization, uncertainty & risk (Monte Carlo analysis), techno-economic analysis, cost- effectiveness, integrated resource planning, custom decision tool development
- 3. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED DER Planning Regulatory Environment DER cost declines & favorable regulatory environments are contributing to rapid growth in these technologies Proactively planning for DERs and understanding their likely impacts is becoming increasingly important to utilities Several states are requiring utilities to integrate DERs into their distribution planning process, such as: • California: via rulemaking 14-08-013 (February 2017) • Nevada: PUCN proposed ruling (July 2018) SB 146 • Minnesota: Docket Number E-002/CI-18-251 (August 2018) Others jurisdictions expected to follow suit as DERs inevitably grow
- 4. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED DER Planning Model
- 5. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED TotalUnitsPurchased UnitsperYear Time Incremental Adoption (Units/Year) Cumulative Adoption (Total Units) Inflection Point Source: Lumidyne Bass diffusion1 and System Dynamics2,3 provide the tools needed to assess distributed solar PV adoption under different scenarios. Stock/Flow Diagram Prototypical S-Shaped Adoption Curve Source: Lumidyne 1. Bass, F. (1969). A New Product Growth for Model Consumer Durables. Management Science, 15(5), 215-227. Retrieved from http://www.jstor.org/stable/2628128 2. Sterman. J (2000). “Business Dynamics: Systems Thinking and Modeling for a Complex World.” McGraw-Hill. New York, NY. Page 332 3. Welch, C. (2017). "Making the Case for Using Analytica for System Dynamics Modeling: A Reference Guide and Comparison with Classical Platforms." Peer reviewed paper presented at the International System Dynamics Conference. Cambridge, MA.. Distributed Solar PV Diffusion Modeling
- 6. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Distributed Solar PV Diffusion Modeling Illustration of a Typical Calibration (Incremental and Cumulative PV System Installs) Source: Lumidyne Historical PV interconnection data calibrated with nonlinear optimization enables robust, accurate spatial disaggregation (e.g., substation or ZIP Code) of forecasts & grid impacts. Spatial (ZIP) Disaggregation Calibration of diffusion parameters to historical time-series adoption data is critical to ensuring forecast accuracy.
- 7. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Battery Storage Mixed-integer programing enables simulating storage economics & hourly impacts for any rate structure Similar diffusion modeling techniques (as PV/EVs) enable adoption forecasting & grid impact calculation Integrated approach accounts for retrofit of installed PV systems & new combined (PV + storage) or storage-only systems Illustrative Optimal Battery Dispatch
- 8. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Electric Vehicles Sales & registration data enable calibrating diffusion & stock turnover by ZIP Code (allocating to circuits, substations as needed) Stock/Flow Diagram Illustration of Calibration EVs/yearTotalEVs Simulated Actual Non-Electric Vehicle Stock Electric Vehicle (EV) Stock EV Retirements Non-EV Retirements Non-EV Additions EV Additions Total Retirements + + EV Market Share (New Sales) EV Economic Utility EV Non-Economic Utility + + + + + + Fraction Familiar with EVs Fraction Unfamiliar with EVs Familiarization Advertising Strength Familiarization through Advertising Viral Strength Familiarization through Viral Effects + + + EV Market Fraction + + <Non-Electric Vehicle Stock> - - + + Electricity Rate for Charging ($/kWh) - Tax Credits & Incentives + + <Fraction Familiar with EVs>
- 9. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Risk & Uncertainty Monte Carlo simulation enables estimating uncertainty & risk of the DER portfolio (including potential correlations) An appropriate metric of the distribution of net impact can be selected (e.g., 90th or 10th %ile), depending on the desired balance between risk and cost. Value at risk (e.g., CVaR90) for the utility is another financial metric Illustrative Distributions of DERs & Net Impact Spatial Distribution of DERs Can be Very Different For Illustration Only. Not Actual AZ Data
- 10. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED SPIDER Model Demonstration
- 11. ©2018-2019. LUMIDYNE CONSULTING LLC. ALL RIGHTS RESERVED Contacts Cory Welch - Founder & President lumidyneconsulting.com cory.welch@lumidyneconsulting.com 720.984.5814 James Milford - Director of Technology lumidyneconsulting.com james.milford@lumidyneconsulting.com 303.548.8847