Part of Grid Network »

The Grid Professionals Group covers electric current from its transmission step down to each customer's home. 

WARNING: SIGN-IN

You need to be a member of Energy Central to access some features and content. Please or register to continue.

Post

How analytics are used as the foundation to drive Grid Modernization

image credit: David PopeImage

This item is part of the Special Issue - 06/2020 - Grid Modernization, click here for more

There are many use cases for using analytics and in particular automated analytics in the form of artificial intelligence(AI) and machine learning (ML) to help drive grid modernization however I would argue there is not one as important as using the proper forecasting inputs that feed into the rest of your Integrated Distribution Planning(IDP) processes. These same forecasts play an important part in your transmission planning processes as well. As a result, these forecasts are the foundation that make up the decisions being made throughout the rest of your planning processes so the better your forecasting process the better the information will be downstream on which decision makers will rely on to make their plans for the future.

Public Utility Commission (PUC) feedback across the US continue to provide feedback on utility submissions that emphasis among other things the need for resiliency in the power grid, improving customer satisfaction, and the overriding theme of the need for granularity in the justification of capital expenditures.  Because PUCs are looking to understand the impact of the shifting generation landscape with the introduction of more distributed energy resources (DERs) in combination with load changes which have been driven by improvements in energy efficiency and/or demand response on the overall electrical infrastructure some of them are moving to include more granular longer-term planning as a requirement for future submissions. We all know there are commonly used IDP tools available which support power flow analysis, grid design, and capital planning, however, the output from these tools/processes are only as good as the forecasts that are used as input or ultimately the foundation for the rest of the decisions that flow from these IDP downstream processes.  Unfortunately, today many of these forecasts do not enable disaggregation of the load, generation components, and economic factors. 

The legacy or traditional approach of using a “simple” model produced by a linear projection over a 5-10 year period is no longer sufficient or really effective enough to deliver the forecasts needed at the granular level in order to instill confidence in utility decision makers and/or the PUCs reviewing long term plans associated with grid modernization.  The challenge of capturing economic or other time series trends in hourly data is generally difficult, but increasingly more complex driven by transfers, DERs, etc.… The good news is this new energy IDP process that requires granularity down to the circuit level forecasts and agility to change or add new underlying factors is possible to put into production today due to advances we have had in technology advances that enable hierarchical forecasts on a massive scale to be run in a timely manner (typically hours) instead of once after weeks of processing.  Let me warn you that while you can use a variety of “tools” to proto-type this type of new process in order to convince yourself and others of the value that will be derived moving this new system from a proof of concept into production is where many utilities struggle to be successful.  Why? Because the environment needed to prepare, process, and use the resulting forecasts on the size of data that is required needs to be a robust IT system designed specifically to do the necessary analytical data preparation, analytical data management, flexible data integration, multivariable hierarchical forecasting for both short-term and long-term forecasting horizons, with built-in governance is only available through the use of an AI/ML platform engineered to support this entire process not just a subset of data typically used to prove out the concept.  While to many in the utility industry this new process may sound like something that will be developed in the future there are actually a few US utilities who are running this type of new system today and as a result they are setting themselves up to be in a better position to meet the needs of the utility of tomorrow today.

Discussions

Dr. Amal Khashab's picture
Dr. Amal Khashab on Jun 21, 2020

The forecast is a forecast, not more. There must be margins from reality. Available technologies nowadays enabling to close these margins. For instance, If a utility discovers a mismatch between available generation and increased demand in a certain area, it will go forward to install a DG unit ( wind/PV + BS or GT) in one week.

Matt Chester's picture
Matt Chester on Jun 22, 2020

The installation would surely take longer than a week, though, wouldn't it? Or are you saying within a week the decision to add more DG would be made?

Dr. Amal Khashab's picture
Dr. Amal Khashab on Jun 22, 2020

The decision can be made in one hour based upon existing data of demand/ supply balance. Nowadays, all RE facilities are modularized. This enables the EPC contractor to manage the whole thing in one week. Remember the cases which happened in the USA during some tornadoes 2017-2019.

David Pope's picture
David Pope on Jun 22, 2020

Dr Amal, Matt,

If that mismatch in demand and generation only happens once or occurs in a pattern wouldn't that influence what decision should be made for both the short term and long term.  While adding a DG might first appear to the "right" decision if you can forecast better into the future your capital planning expenditures long term may be better off spent adding a different DG or multiple DGs in different places.

David

David Pope's picture
David Pope on Jun 22, 2020

Dr. Amal,

While I agree a forecast is a forecast, some forecasts are better or more useful than others.  A granular forecast that can break out different levels of the grid and incorporate new data quickly and be re-run fast provides a better more flexible process than one that relies on a forecast that is 1 day, 1 week, 1 month, or more old.

Regards,

David

David Pope's picture

Thank David for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »