Smart meters as a demand response enabler

Time based rates enabled by investments in advanced metering infrastructure are shown to elicit peak demand energy reduction by residential customers.
Published: Fri 17 Feb 2017

Since the early days of smart meters, when their deployment was started primarily as a means to improve meter reading, the impacts they have had – and are having – on the electricity system have grown significantly.

Engerati has identified grid intelligence and grid flexibility as among the main impact areas of smart meters, which will be investigated in depth in future articles.

But arguably the most important is what has always been cited as a key customer benefit – namely their potential to enable the customer to save energy and therefore money, with the smart meter in effect becoming a tool to enable demand response.

Since the early days of deployment at the residential level, debate has raged long and hard about just how effective the smart meters are proving in leading to savings and whether there is a short term ‘novelty effect’ or that any savings can be sustained in the longer term.

To investigate this further we turn to the country that has led the smart meter charge, the United States – now with a national penetration over 40%, according to FERC’s latest assessment.

In particular, a series of 11 consumer behaviour studies conducted by 10 utilities with support from the Smart Grid Investment Grant (SGIG) programme, the final results of which have recently been released.

Consumer behaviour studies

The 10 utilities are Cleveland Electric Illuminating Company, DTE Energy, Green Mountain Power, Lakeland Electric, Marblehead Municipal Light Department, Minnesota Power, NV Energy, Oklahoma Gas and Electric (OG&E), Sacramento Municipal Utility District (SMUD) and Vermont Electric Cooperative.

Together these were considered a “generally representative group of utility types, sizes and regions of the country”.

The goals involved applying randomised and controlled experimental designs for estimating customer responses in order to advance understanding of time-based rates and customer systems, and provide new information for improving programme designs, implementation strategies and evaluations.

Each of the utilities evaluated at least one of four types of time-based rate programmes: critical peak pricing (CPP), critical peak rebates (CPR), time-of-use (TOU) pricing, and variable peak pricing (VPP), In addition to rates, the utilities evaluated a variety of non-rate elements, including in-home displays (IHDs), programmable thermostats (PCTs) and education.

Lastly, all the utilities employed an opt-in (voluntary) recruitment approach to their studies, while two augmented that effort with a separate opt-out approach (where customers are automatically defaulted on time-based rates).

Consumer behaviour findings

The findings are briefly summarised as follows (with more detail in the table):

Summary of major findings of smart meter consumer behaviour studies

• Opt-out recruitment approaches resulted in higher enrollment rates (92% vs. 15%) but generally lower peak demand reductions (6% vs 12% for TOU and 13% vs 23% for CPP) than voluntary enrollment methods. However, retention rates were about the same for both (90% vs 87%).

• Retention rates and demand reductions were higher for CPR (89%, 21% respectively) than for CPP (80%, 11%), However, when PCTs were available, the differences in average peak demand reductions between CPP and CPR were largely eliminated.

• Free IHD offers did not make a substantial difference for enrollment and retention rates. SMUD’s peak demand reduction estimates were larger with IHDs (2-3%), but this can be attributed to pre-treatment differences between the two groups so there was not a measured IHD effect on reductions of peak demand.

• Free PCT offers did not make a major difference for retention, but peak demand reductions were substantially higher when a PCT was present (22-45% vs -1 to 40% without).

• Estimated peak period demand reductions from customers exposed to TOU rate ranged from -1% to 29%, with an average 15%. Demand reduction increased with increasing peak to off peak price ratio (6% with ratio less than 2:1 vs 18% with ratio greater than 4:1). When CPP/CPR was overlaid on the TOU rate, the average event peak demand reduction was 27% when averaged over all of the treatments. When PCTs were available, the average event peak demand reduction was 34% vs 24% when not available.

Impact of smart meters on DR

These findings indicate the clear benefit of time-based rates in enabling a demand response from customers. Notable is the added benefit of a PCT compared with an IHD.

Among the outcomes of the projects is that three of the 10 utilities allowed participants to continue taking service under the rates after their study was completed.

Four of the utilities chose to extend an offer of the rates tested in their study to the broader population of residential customers. Specifically, OG&E has enrolled approximately 116,000 residential customers (representing approximately 18% of the residential population) on its SmartHours programme, 100,000 (86%) of which are taking service on the variable peak pricing rate tested in the study, and are achieving 147MW of peak demand reduction.

This voluntary SmartHours programme includes the offer of a free PCT, which 90% of customers have taken. SMUD has chosen to make the TOU rate it tested the default for all of its residential customers, starting in 2018.

More broadly, the California Public Utility Commission ordered  the state’s investor-owned utilities to make TOU the default for residential customers, citing heavily the very positive results SMUD achieved as grounds for this decision.

Related Webinar