Interview with Alan Ross - "A Reliable Grid is a Smart Grid"

Alan RossAlan Ross is the Vice President of Reliability for S D Myers. He is responsible for developing and executing long term strategies and next generation leadership for all operating units.

He is a credentialed reliability professional with both the CMRP and CRL certifications and Is a member of the IEEE Reliability Society. Alan Is the Chairman of the Smart Grid working group for SMRP and the Electrical Power Reliability Summit and on the Planning Committee and Keynote speaker for the Comet Conference.

He is a dynamic and frequent presenter or keynote speaker at NETA, Comet, EPRS, SMRP Conference and Symposium, Marcon, Reliability Conference, AIST, IMC and numerous Muni/Coop regional organizations.

Alan has published frequently in AIST Journal, Plant Engineering, Solutions Magazine, Uptime Magazine and on the blog Transformer Reliability, and numerous white papers on the adoption of new technology, reliability and leadership.

In this interview, Alan answers questions from his webinar, "A Reliable Grid is a Smart Grid", originally presented on Dec 20th, 2018. For more details regarding these questions, please view his webinar on-demand on the IEEE SG Resource Center.

Can storage be used to reduce the amount of cycling inflicted on older fossil-fuel generators?

Yes, storage can help certainly, but the volume needed to offset the cycling issues are not yet attainable. I heard recently at a conference that the answer will be “flow” batteries, but that has been a promised future that has not yet happened in enough volume to overcome cycling issues.

Can you give some specific examples of coal plants that are being adversely cycled because of DER?

This year, there will be a large number of coal plants either shut down or converted. Some are being mothballed, so they technically still exist but are not generating. Why? Is it all about issues with coal or is more about the money. There are far too many reasons associated with adverse cycling, than just the one I spoke about. It is a complex issue.

Transformer bathtub curve mentioned on slide 12 shows 10 to 20 year interval with constant failure rate. Does this mean, no ageing takes place of paper, oil or other components? Also what about the impacts of maintenance failures like failure to rectify a leak. Will it change the curve?

Yes there is degradation of the paper, but that does not count as a failure. What the curve tells us though is that during this period of time, when the rate seems constant it is actually “random” in nature. Taking care of the oil, which takes care of the paper will extend the life for the most part into the ranges well beyond 20 years. That’s why we have an average age of power transformers in the US of over 35 years.

Technical aspects are still needed, however, how does IEEE (and tech stakeholders) help change the utility business model so IOUs (which manage the largest amount of the grid) invest into 'smartening' up their Distribution Grid/Transmission Assets?

There is a stark difference between transmission assets and distribution assets and the “smartening” up you mention is and continues to happen, mostly starting with transmission. When we talk about distribution, it gets much more complicated with coops and muni’s providing a lot of the service in the US. It’s here that we see the greatest need for “smartening” but it’s also here that we see more disparate solutions and too little investment to make it truly a smart grid. Add to that the problem with DER and the reverse flow of energy into the grid and we can see where the vulnerabilities are. Distributed Automation (DA) is thought to be the answer, but that is not certain yet. IEEE is working ahead of the curve on most if not all of these issues.

So are you saying that we as engineers cannot solve the problems of Smart Grid but AI will?

We as engineers are part of the AI equation. Where does the information flow for AI come from? Usually engineers over decades and who will use AI? Engineers and Data Scientists. Machine Learning (ML) is likely a more practical application of algorithms and engineering coming together to solve problems and build a smart grid.

What is AMI?

Advanced Metering Infrastructure. It’s Smart meters at the consumer level and they are getting smarter all the time. Check it out here.

Does reliability depend on interdependency? If yes, which factors needs to be considered to do interdependency analysis?

Oooh! Best question on here and one that cannot be answered that simply, but YES! System reliability is much more than asset reliability and it is a particular area I am personally passionate about. The synergistic reliability (Resilience, redundancy) of a system is more than the sum of asset reliability of its parts. There is an interdependence in the grid that must be considered. One simple example: We add underground cabling (UG) into an existing above ground (AG) loop, let’s say into a suburban neighborhood, because new home owners don’t want to see ugly poles and lines or they don’t want their 100 year old oak tress cut into a “V” to make way for lines.

Where do you place reclosers on the loop? What impact will this new UG have? It’s not simple and it’s not cheap. There are so many more issues that impact interdependency when you consider the impact DER alone is having on the grid, especially the Grid Edge. Contact me, I’d love to discuss more.

Is it possible that the increased complexity of the smart grid could lead to a less reliable grid?

Absolutely and the question above is the reason why, but I do know that NERC, EPRI, IEEE and IEC are working on a lot of these complexities to increase reliability, even as complexity increases. And the impact of microgrids could in fact remove some of the unreliability by simplifying the generation and distribution within a closed loop. Of course, there will be the need to back that loop up to a Smart Grid, which does add more complexity once again.

In terms of machine learning techniques for PV forecasting, which technique is the best? Which is easiest?

Sorry, but this one is above my pay grade since I am not a Data Scientist.

Of the many optimization methods in data science which one is the best method and why?

Again, see above?

To view past interviews, please visit the IEEE Smart Grid Resource Center.