A day in the life of an Energy Intelligence Manager
Our colleagues are at the heart of the ESO, working hard to operate a safe and reliable electricity system, whilst supporting the transition to a greener, and more efficient system for future generations.
James Kelloway, Energy Intelligence Manager, gives an insight into his time at the ESO
What is your role and when did you join?
I’ve been with National Grid since 2013.
I have a pretty interesting career path! When I was a teenager, I studied classical music. I’m actually a trained conductor and play the Tuba and also play the Piano badly!
Then I went to HR and Payroll systems development, commercial radio and a lot of IT and data warehousing work.
My varied background is both good and bad - you see a lot of problems and solutions that many people don’t see when you’ve been in a lot of industries. But then I balance that out with colleagues who have been here for many years and have deep routed industry knowledge.
Can you tell us about your current role?
I head up the ESO lab which focusses on the long-term view. What will change? What will we do? What do we need to test? It’s great that I get to work with top tier data scientists to look at new problems and challenges.
We look at three pillars:
1. Pure R & D
We test! Most tests will fail, and we learn. But some change the world!
An example of this is the science behind carbon intensity. We’ve seen around 19m hits using that data service in a recent month. To give you an idea of how big that is, the NHS website has roughly 50m hits a month!
The data tells people how clean electricity is now and how clean it will be. Our carbon intensity app works off that. I’m proud of the app and we worked with WWF, Environmental Defence Fund and Oxford University on the data behind it. This data also allows us to measure how close we are to our zero carbon ambitions. I’m passionate about making the world a greener and better place, and we’re making fantastic progress at the ESO.
The rate of change is bigger than ever before, and data quantity is higher than ever. We have to be pro-active because of the zero carbon ambitions and amount of micro generation on the system.
Our data (below) shows our progress to zero carbon electricity generation. We’re doing well but the closer we get the harder it becomes.
2. Proof of concept
This is when we think this might work but we’ll try it and see if it works. We’re currently looking at electralink data. It will help our embedded forecasting which will mean less uncertainty for our control room and lower costs for consumers.
Some estimates are between 30-40% embedded so the rest is managed by balancing. Some days, this is much higher. The embedded gen is driven by two key things – weather and cost.
If we understand those and can build them in to our forecasts, it means we won’t need to dip into our reserves as much. It’ll also potentially mean lower carbon operation.
3. Capex delivery
When we have an established product we feed in and produce applied science with the delivery teams. Machine Learning (part of AI) is something my team is focussing heavily on.
We commissioned a first of its type super-computer using the term!) a couple of years back to process the vast amounts of data needed. Back in the 1970s we had the biggest super computers outside of the military and we’re pushing that forward again.
I believe our new computers have the biggest processing capability within National Grid – for the techies out there, the GPU engines (each machine has 10) has a max memory throughput of around 900GB/s (a standard grid laptop is about 4GB!) with tensor performance of 125 TFLOPs! That’s fast, very, very fast.
But we need it for the tech to work with the large data quantities so we can predict within operational timescales. What I really like about this setup through is the power consumption... each super computer runs on only slightly more electricity than the average kettle. Pretty awesome for that amount of power.
Does your love of technology flow through to your personal life?
I’m on my 3rd electric car and my house is totally smart. Three of the five of us in the lab have an EV so we’re probably the highest proportion team with EVs in the business!
In my role I see a lot of the data, and I visit many universities to speak to experts. I know had bad petrol and particularly diesel is and can no longer justify renting a Mustang on holiday! Fortunately, Ford are making an electric version...
The tech also saves me money. I have a tariff where it’s super cheap for four hours at night so that’s when I charge my car. Works out about 1-2p per mile for zero tail pipe emissions. Can’t say much better than that.
Long term technology is going to help consumers a lot. For example, you’ll probably keep your car plugged in at home and it’ll charge when electricity is cheap or even negative.
Energy companies may well become IT companies which buy and sell energy. Some already use APIs with their flexible tariffs and we are going to see a lot more of the tech upscaling as the smart energy market matures.
How have you found the restrictions due to the COVID-19 pandemic?
I’ve found it tougher personally than professionally. I’m a single dad and it was difficult not having the kids in school. That said it’s great spending time with them, it’s just tough getting the balance right.
My commute is super short (about seven minutes on a bad day), so I don’t have the same time saving that many do when working from home all the time. I do think this year will change flexible working for the better and offices will likely become more like productivity hubs rather than bases in the future.
Luckily, I have a great team who are used to getting on with work and love doing it. They’re bought into it and I’ve seen no dip in performance. I have to give credit to them for adapting so well and continuing to deliver excellence that drive us toward the 2025 goal of zero carbon operation.