Way forward for electric utilities – Better data, better analytics and the...

Way forward for electric utilities – Better data, better analytics and the ability to predict

SHARE
Robert Sarfi, Managing Partner, Boreas Group LLC

In conversation with Robert Sarfi, Managing Partner, Boreas Group LLC, Geoff Zeiss, Editor- Building & Energy delves deeper to understand the importance of data management and the challenges electric utilities face.

So far how has been your experience working with the utility sector?

I am working with electric utilities since 1993 and I am involved in strategic technology planning, deployment, cost justification and benefit validation. I try figuring out how a utility company can best leverage technology to drive business value. This is, in fact, the core of my business since last decade. We develop concepts and visions for modern utility and build plans to put utility on a roadmap to achieve those plans. In the last five years, we have done national level smart grid plans, primarily in West Africa and South East Asia.

What are the key aspects of smart grid if we talk about utilities in Americas and internationally?

Utilities across the world face a few challenges — safety, reliability and cost of service are the top three. These are symptoms and the solution is utility-specific. You have to look at their issues, their priorities and then figure out the best possible solution.

As far as smart grid is concerned, several countries have only one utility company. What are the major challenges they face?

Most of these are supported or implemented by the Ministry of Energy. They may have one or two utilities supervised by the Ministry. The challenge is – these may include existing infrastructure issues, adequacy of supply issues and viability issues. And then, we have explosive growth potential. Imagine a North American or Western European utility with all their problems exaggerated to 15 times. There is so much higher impact than we are used to see.

What is at the back of your vision when you strategise a utility? Do you find clarity in what utilities consider as their priorities?

They have priorities, conflicting ones at that. That’s where the real challenge comes in when we are building a vision. I tell people that the new regulatory requirements or an external event is good for a utility because it calls for action. You have to work with them to create a balance.

Data quality is the biggest challenge utilities are facing. How have utilities changed their business processes to optimise data quality?

There are a couple of utilities, which truly believe in the power of data, but there is a cost associated with that data. They engage in a data cleansing activity and find ways to not corrupt the data.

Utilities, at times, spend on data cleansing operations and then they neglect data management. Your take?

You need predictive modelling to do this — a detailed model that predicts minute, hour, day at least two days ahead. For such accurate modelling, we need good connectivity model and better understanding of our loads than ever before. Predictive modelling is able to predict when a transformer is going to fail based on its load pattern.

Have you come across utilities that have realised how valuable data is?

Data quality and predictive modelling are two key aspects of smart grid deployment and both involve data. Right now those who are doing a really nice job are doing it using focussed applications. A lot is being done in the transformer load calculation too. Those who have used these pointed applications have realised its power and, in fact, been encouraged to engage in more such projects. However, they also realise that in order to do this, they need to clean up their data. In lot of cases, they fix up the data, but many a times data is stored in silos so it is difficult to correlate and connect data. So, they end up doing data mash-up and try to visualise it, which is almost impossible.

Utilities are trying to bring data together in some way,and now see the importance of geospatial technology. How is GIS is becoming a foundational technology compared to being a tactical technology?
Every system has a concept of sub-stationing. So, every operational engineering based system will have sub-stationing. When you try to do analytics using large volumes of data, if the sub-station doesn’t match, you are in trouble. It is just one example, on the surface you think how difficult it is to match all sub-stations across systems. It is much harder than you think, especially in a big utility company.

A study says by 2018, cloud services will make half of the IT portfolio of over 60% of utilities. Do you see that happening considering that SMEs do not have either geospatial or IT capabilities?

I really believe in cloud based solutions. The issue I have with cloud based solutions right now is the cost of infrastructure associated with it. The primary motivation for cloud solution is commercial. These utilities — small to large — feel the cost is too high that they feel they can buy a supercomputer and employ researchers in the service cost offered.

The traditional GIS model is what is called infinite time model. You collect the data, analyse it and it takes a long time to submit the report. In utility space where we have spatial data coming from umpteen sensors, this needs to be done in microseconds. In this new paradigm, how do you see geospatial being integrated with other processes? Will today’s LBS and GIS vendors be able to achieve this or we will see a whole new bunch of companies, either start-ups or companies like Trimble, bringing a lot of technology solutions in one basket?

A few players will be there for a long time. GIS of the future might consider location, involve real time data transfer and analytics; and might look at both past and the future. Past is really important when you are talking about predicting the future. Future is exciting with new entrants joining the game and there’s going to be three dimensional or may be four dimensional spatial components. Part of this is going to be market-driven. The tough part for most utilities is that it is regulated and that doesn’t necessarily promote a culture where innovation can exist freely. I see that in some Asian markets where business models are changing owing to supply scarcity.

Business models keep on evolving in North America utility market, which has about 3000 utility companies. States are working on new business models for utilities. Fundamentally, they are de-coupling their revenue from selling power. Do you see this happening in Asia and other parts of the world?
I feel, in the coming decade we will see a completely different business model. We have to challenge ourselves so that we are not part of the problem. The PVs and solar city models that are available in the market today are now posing a challenge to the traditional models. Fundamentally, for utilities, there should be better data, better analytics and the ability to predict things including climate change.

So, GIS is absolutely fundamental since it brings in all the data from this siloed applications together?

There won’t be change overtime. I am not sure how our business models will look like after two decades. But I know that if I can control my data, I can do the right analysis and respond and articulate well to my stakeholders. It’s about how you can be in control of your business by controlling the right data.

Talking about the new labour force, what kind of training they require to handle the evolving utility practices and processes?
My generation and the generations before me were building the infrastructure that we have today. We didn’t have real-time visibility that people now do. What excites me is the fact that youngsters today are smart and analytical. They bring skills that we never had access to I hope they use the foundation we provided them to the maximum and solve problems that we never thought we could.