Publication - Working Paper
Theft/non-payment for electricity services is a concern for the utilities in the Kyrgyz Republic, as with electricity utilities in many developing countries. Low payment rates jeopardise the financial stability of the utilities and their ability to carry-out maintenance activities in the short run, as well as limit their ability to make longer run investments necessary to meet growing electricity demand
In collaboration with an electricity utility in the Kyrgyz Republic, we are implementing a randomised study on the impacts of and benefits from smart meter installation amongst residential consumers. Specifically, we are measuring the impacts of household smart meters on electricity utility bill collection, distribution losses (technical and non-technical), and quality of electricity services (power outages and voltage spikes). Most of the existing research on smart meter technologies has focused on their role in implementing time-of-use (and other forms of dynamic) pricing and specifically focused on developed countries. We believe there are potentially great benefits from installing the smart meter technology in developing countries, but with a very different role and intent than in the United States or other developed countries.
To carry out this study, we are randomising the roll-out of a new smart meter technology at households in Kyrgyzstan starting in December 2016. By also installing smart meters at neighbourhood transformers, we will measure electricity losses as the differences between the transformer-level measurements of consumption and the aggregated household-level measurements of consumption. Engineering algorithms will allow us to distinguish between technical and non-technical losses. In addition, we will conduct household surveys to measure both the short-run and long-run household response to smart meters on outcomes such as electricity consumption, energy efficiency investments, appliance and other asset ownership, and small business activity, amongst others.
By measuring all of the outcomes above, we will identify the channel(s) through which smart meters impact household behaviours (for example, through improved reliability or increased enforcement of bill payment). Utilising data on changes in electricity service reliability and household asset purchases in a discrete-continuous choice model, we can estimate the demand for improvements in electricity reliability.