Presentation Details
| Resilience Metrics Framework for Solar Photovoltaics Bonnie Powell, James Elsworth. National Laboratory of the Rockies (NLR), Golden, CO, USA |
Abstract
Solar photovoltaics (PV) systems can be resilient to extreme weather events; however, they are sometimes damaged during those events. Storm hardening measures and system attributes can significantly reduce the likelihood of damage to PV systems after an event. This work proposes a resilience metrics framework for quantifying, comparing, and predicting the resilience of PV systems. The framework includes attribute metrics independent of the event and performance metrics post-event. Attribute metrics are weighted and aggregated into an attributes resilience score. Performance metrics are normalized by comparing observed performance to the expected performance for a system experiencing a storm of similar intensity, weighted, and aggregated into a performance resilience score. Key to this framework is the expected damage of PV systems for storms of varying intensities. This work presents damage function curves from actual PV systems that experienced hail or windstorms to estimate expected damage. Finally, the presented framework is applied to three case studies of PV systems that experienced damage from hail or windstorms, and the observed damages are compared to expected damages predicted by the damage function curves. Although additional data collection is required to establish weights for the resilience scores as part of the presented framework, this work 1) presents a methodology for quantifying and predicting resilience of PV systems to storms and 2) includes damage function curves to add to the body of literature on establishing expected PV damage from storms. The damage function curves predict less module damage from hail than found in laboratory testing, perhaps due to variations in PV system attributes. PV system attributes—in addition to storm attributes—influence the extent of PV damage, and additional damage data with accompanying PV system attribute granularity is needed in future research.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.
No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.