Can Developments in Smart Grid Analytics Help Us with How We View and Use Air Compliance Data in the Future?
Overview
This blog is part of our continuing series on the data side of air compliance looking at changes, trends and technology that can improve the availability, usability and accuracy of air emissions compliance reporting. In our previous discussion we talked about hosting emissions data in the cloud and the feasibility of using edge technology to enhance or replace traditional data acquisition and reporting systems (see March 1, 2019 post for details). Today we are looking outside to other industries to learn how they are using smart device technology and data analytics to change the way business is done. Our goal is to gain insights that will assist the emissions industry to maximize technical advances in data collection that will improve compliance reliability and operational excellence.
The basis for today’s report and commentary come from a recent webinar I attended titled Growth Opportunities for Optimizing Electric Grid through Data Analytics. This was hosted by BrightTALK on April 11, 2019 presented by Farah Saeed, Research Director, Digital Grids with Frost & Sullivan. The views and comments in this synopsis are wholly my interpretation of what I observed during the webinar and may not necessarily be 100% in line with how Frost & Sullivan researchers see it. I encourage all my readers to check out the presentation firsthand on the BrightTALK website or by the following link found at the end of this blog.
The referenced presentation was a big picture, high level view of the world focusing on market perspective, trend observations, change predictions and growing acceptance of data analytics in the evolving utility business model. The move to more effective data analytics is one part maximizing competitive advantages through technology and one part balancing modern electric generation with consumer demands for more reliable services. Successful application of data analytics is about business survival. As utilities invest in more and more smart technology to help manage the grid, they must learn to effectively use data pulled from many sources to allow them to do it profitably. The evidence is trending towards integrated analytics and the underlying message is adapt or risk failure down the road.
What can the air compliance community learn from this utility model? Hopefully we will see a common path leveraging similar smart device technology, data collection devices and software tools that will easily translate to the emissions reporting environment. Also, we will learn whether the same challenges and barriers of entry exist for both or if the emissions monitoring market has its own, uniquely different set of resistance rules. Time will tell the future, but in the mean time we will use every resource available to keep our finger on the pulse and to stay ahead of the technology curve. For the sake of efficiency and not wanting to write a novel, we will present the utility side of the equation today and close with our thoughts on how data analytics might fit into the emissions monitoring world.
The Utility Problem
There are measurable trends in the utility sector that define how our electricity infrastructure is changing and its impact on how utilities do business. These trends include increased acceptance of electric cars, smart metering, energy storage (batteries/fuel cells), smart grid functionality, renewable energy expansions and IoT technology additions to transmission networks. Utilities must find ways to adapt their business model to handle changing market conditions, price reductions for services, availability/integration/cost of new technology into the system and customer acceptance of the way that business will be done in the future (change management). The challenges overview for providers include:
· How to keep pace with trends and manage through the introduction of new technology, with little or no interruption of service.
· How to balance need with changes in power generation (meeting peak demand in conjunction with integration of renewable energy sources into the supply gird).
· How to use IoT, Artificial Intelligence (AI) and machine learning to improve/guarantee service reliability.
The primary data paths that utilities would use to improve analytics are sourced from: transmission and distribution networks, smart meter infrastructure, customer service/satisfaction systems and directly from customer energy consumption patterns. The problem with the shift in how business is done is being able to “read” and make sense of all the data coming from so many different sources. The data needed to create the analytics model is not fully available on a common platform, giving us a lot of data with no integrated or logical result. Some of this data is easily accessible, available electronically, some of it is still manual entry and most of it drawn from unconnected sources. Organizing it and making it meaningful is the challenge.
Outcome Based Analytics (OBA)
Analytics as a solution starts with effective usage of the available data. Understanding what we can do now and where we need to go later is the first step towards success. The development and use of outcome-based analytics to integrate the available data components of generation to turn raw, disconnected data into useful data is moving utilities in the right direction. Power generation is changing through the addition of more renewables sources (i.e. impact of solar or wind on the gird vs base load units of the past) and by bringing smart devices into the supply/demand management system. As the developing analytics get more and more refined, utilities will need to turn consumers into prosumers by engaging them in the process so that they adopt the changes willingly and remain engaged throughout the change cycles. With this approach power producers can balance new generation sources and consumer patterns using grid analytics to manage demand more efficiently. Time and money will drive improvements as we manage through until the technology transition goes mainstream.
Based on the current trends and technology infrastructure that is already in place the smart grid development market is estimated at over $50B globally. Expectations are that it will continue to grow from here as data analytic efficiencies improve. Utilities are generating massive amounts of data from the billions of dollars invested in transmission & distribution devices, smart meters and customer service programs, but are currently leveraging less than 5% of this data for business improvement purposes. Slow adoption is caused by high upfront costs, low confidence in the sophisticated analytical tools (trusting AI vs human generated data) and concerns about consumer acceptance of the new model. Trends are improving but the market won’t fully reveal itself until about 2025 (estimated) according to the experts. Still with a starting point of $50B the future seems bright for smart grid data analytics and the software providers that develop them.
Outcome-Based Analytics (OBA) is seen as the future but for now, buy-in is slow going. The available data is not always compatible with traditional data analytic tools which impacts investment decisions. To drive the smart grid market in the right direction, OBA providers must gain widespread, committed adoption through the development of advanced data algorithms that can provide reliable and believable outcomes. The industry looks at four layers of outcome analytics to help define software models and to build confidence in the available tools. The four layers as listed in the presentation are:
Diagnostic: the ability to detect or describe patterns or attributes within a data group
Descriptive: the ability to describe simple patterns or characteristics within a data group
Prescriptive: the ability to prescribe a pattern and deliver analytic outcomes based on AI
Predictive: the ability to predict a pattern or data group impact as well as evaluate which intervention contributes the most change
Frost and Sullivan predicts a shift from core diagnostic and descriptive data sets to more prescriptive and predictive inputs going forward. Simply put, moving data analytics confidence from the old school human interfaces to the AI basis of prescriptive/predictive outcome analysis. Investment in network infrastructure, growing smart meter usage and adoption of edge devices for data collection would all be needed to complete the baseline shift supporting market advancement. A significant change in data sourcing is expected as early as 2021-2022 due to the availability of 5G based connectivity and subscription-based cloud solutions. With the growing reliability of 5G communications, analytical software will be able to convert data real time using the statistical models built from prescriptive/predictive inputs to flag inefficiencies and provide the ability to predict events before they happen. Under this belief, AI enabled analytics can deliver a wholistic decision making tool improving the quality of the utility business model while gaining consumer acceptance/confidence in the service changes.
Core Changes Driving Utilities Forward
There are three main data sourcing areas that are being addressed with today’s improvement investments: Transmission & Distribution Analytics, Meter Analytics and Customer Service Analytics. Software and technology aimed at improving these data collection points along with increased reliability of prescriptive and predictive algorithms is expected to grow rapidly over the next 5-7 years. Each major area has a different path to follow, leading to the overall improvement of OBA products in the marketplace and utility confidence in the data they provide. The key adjustments for each area and the associated drivers for improvement are summarized as follows:
Transmission & Distribution (TD):
1. Significant drop in the cost of sensors and intelligent devices with further reductions expected will help drive investment in improvements.
2. New investments are driven by regulatory requirements for minimizing downtime and energy loss. Managing energy demand is critical to a utility’s success.
3. Retrofitting of distribution networks adding a physical layer for intelligent devices to improve energy delivery and grid operations optimization.
Metering Analytics:
1. The growing installed base of smart meters has significantly improved quality of usage data and in defining peak demand periods which leads to better grid management.
2. Approximately 575M devices were deployed globally by end of 2018 with an additional 865M expected by 2023.
3. Average cost has dropped about 25% since they first started becoming mainstream, which helps drive utility’s commitment to use smart meters.
Customer Service Analytics:
1. Customer analytics are critical to influencing program strategy, marketing approach and value add services.
2. Foundation is predominately built on CRM expertise (think SAP or IBM) leveraging experience to build customized solutions for energy utilities. Helps move from antiquated customer service models to software-based systems.
3. Development of energy management programs focused on balancing load and supply, reducing energy bills and improving operations efficiency (moving consumers to prosumers).
The suppliers serving this market have the choice of being a “section provider” or an integrated provider serving all three OBA growth paths. Frost and Sullivan has broken the competitive market for smart grid OBA products into 3-Supplier Tiers as follows:
Tier I SAP, Oracle, IBM, Siemens, ABB, GE, Sensus
Tier II Landis & Gyr, C3 Power, Tendril, Honeywell
Tier III Nokia, Uptake Technologies, Autogrid, Tantalus
Obviously, the Tier I players have the advantage leveraging current positions in other areas of utility management to branch into the OBA sector. They have the fire power and money to meet changing demands and keep pace with development needs. My sense of it is that after Tier I the other Tiers are focusing on serving smaller subsets and sectors as opposed to diving into a fully integrated solution. That said, this is a growing field and if predictions are accurate will be very profitable in the very near future (less than 10-years).
Smart Grid & Data Analytics Conclusions
1. Most utilities are moving towards improving their data analytics through advancing their prescriptive and predictive analytical models and investing in infrastructure to improve data quality.
2. More smart devices and development of meaningful data analysis algorithms supporting statistical modeling will come on line over the next 5-7 years, further refining the quality of and confidence in Outcome-Based Analytics.
3. The industry will continue to see variations of business models that utilities can use to improve performance. These variations will come from inside and outside of the industry as technology and communications improve globally (i.e. 5G availability).
4. The industry will see the value of changing their business model from being just a power producer based on generation assets to a service provider dependent on outcome-based analytics.
5. Barriers of entry for providers will be reduced with widespread acceptance of AI based technologies, lower installation costs, improvements in communication infrastructure and data security confidence.
Key Takeaways for Air Compliance
Air compliance data collection and reporting based on smart technology and data analytics is following a similar if not slower path to acceptance as the utility grid sector. The challenges in making the technology investments, shoring up networking infrastructure and merging data onto common platforms are the same for everyone. One thing that was not addressed in any detail, in the referenced presentation, was the issues and concerns surrounding utility confidence in cloud-based services and data security. It was mentioned but not presented as a key focal point. This has traditionally been one of the main barriers of entry for the regulatory market. If you don’t have the data and can’t report it accurately then you can’t run your facility without being in direct violation of your operating permit. Concerns about security and data reliability of cloud storage based solutions seem to stunt the shift in this sector more so than in utility example. This is not an insurmountable concern but certainly one that impedes short-term progress.
That said, there were many positives we can take away from this review, including but not limited to the following:
1. Everything related to OBA and the underlying layers described in this presentation can be adapted to fit the environmental market. As previously reported in our March 1st blog, edge devices and cloud-based compliance is already in play and will grow in acceptance over time.
2. OBA analytics will find a place in environmental monitoring. To some degree companies like VIM Technologies, Inc. are already applying these techniques to predict emission outcomes based off source level, plant operating data brought into their stand-alone data acquisition systems. Power generation relies heavily on the demonstration of compliance. If you can’t comply then you can’t run. If you can’t run, then a utility can’t generate profit. OBA can only make this a better, more reliable model.
3. Prescriptive and predictive based algorithms and the follow-on statistical modeling developed for utilities will speed development in other sectors and strengthen confidence for the regulatory community. Utilities are one of the largest industries affected by the air emissions regulations. The transition from smart grid analytics to regulatory compliance applicability may simply be a matter of getting both sides of the operation in the same room at the same time to connect the dots.
4. Advances in data sourcing and the creation of common collection points in the utility smart grid sector is another area that will benefit air compliance. Over our 30-years in the air emissions markets we have seen a significant move towards using plant data to enhance DAS usability allowing the compliance system to service several levels of the regulatory requirements, beyond just monitoring the smokestack. It can only get better from here.
5. The Tier I players in the smart utility market are the same players that already occupy space in all industries affected by air emission regulations. For this reason, they will also be the ones instrumental in moving environmental engineers from traditional compliance data acquisition and reporting systems towards advanced software solutions. We don’t see these Tier I companies being the providers of the compliance solutions per se, but rather the influencers that will drive change. It will be the job of the DAS software providers to step up and make it happen.
Whew! A meaty topic to be sure but one that needs airing out more and more these days. Thanks for hanging in there and letting us share our thoughts with you. As always, we welcome feedback and your views both supporting and opposing. Afterall it is all about the dialog. Have a great day.
To learn about how utilities are adapting to changes in grid optimization please refer to the following link for details.
Webinar: Growth Opportunities for Optimizing Electric Grid through Data Analytics
Hosted by: BrightTalk
Date: April 11, 2019
Presented by: Frost & Sullivan – Digital Grid Division