AI, Complexity, and Cross-Industry Collaborations: Key Trends at T&D World Conference

The electric utility industry is at the forefront of infrastructure related optimization efforts when it comes to addressing the complexities, cross-industry issues, and regulatory challenges associated with clean energy, e.g., for freight transportation, or for heat to replace fossil fuels in industrial processes, or for clean energy for feedstocks in other applications.

The T&D World Live Conference in Atlanta, Georgia, this past October had compelling sessions across five tracks of importance, including: AI and Digitalization, Distributed Energy Resource Integration, Electrification and eMobility, Grid Resiliency and Black Sky Hazards, and the Future Transmission & Distribution Grid.

Largest U.S. Utility Mutual Aid Assistance: Hurricane Helene

Electric service restoration efforts in response to Hurricane Helene are the largest utility industry mutual aid assistance mobilization in U.S. history. Tens of thousands of utility crews from all over the U.S. have been participating in restoration efforts. Τhe scale and depth of the storm’s destructiveness in some areas of the Southeastern U.S. is requiring a full rebuild, rather than mere repair, for portions of impacted T&D systems. While this reduced utility attendance somewhat at the conference, nonetheless, personnel from utilities and their solution and service provider communities participated in numerous productive meetings and panel sessions.

Greater Collaborations are Underway and Growing

Heroic utility line workers’ collaborative spirit was praised at the conference. Cross-industry collaborations at the conference are also praiseworthy.

Utilities’ long-standing traditions of mutual aid point to key wider trends, because much greater cross-collaboration was evident at the conference in other ways that span across industries who share risks associated with infrastructure and supply chains, including those associated with energy and with transportation, all of which rely directly and indirectly on electric utility service.

Growth areas are impactful, for example, for the oil and gas, chemical, mining and metals, and diverse manufacturing industries, given the economic value and other benefits associated with electrification and replacement of highly GHG emissive processes with cleaner energy sources ranging from renewable electric power to hydrogen to other sources.

To achieve success, innovations driven by competition are an important factor. But it will be insufficient. We also need greater collaboration, and a vision on the part of industry that can appropriately position policymakers to enable the infrastructure resources we all share.

Challenges Addressed vs. New Challenges

The integration of wind and solar on the electric grid is an enormous success. At a presentation by the California Independent System Operator (CAISO), the fact that the drop in solar output during a partial eclipse last year was handed routinely amazed this analyst, when noting that the one hour drop in question represents roughly the same amount of electric generating capacity required to support the five borough and surrounding suburbs of NYC in Con Edison’s territory on a typical day. Con Edison’s current peak is at around 13,000 MW, and when I worked there as an engineer in the mid 1980’s, it was a big deal when peak demand of 10,000 MW was exceeded for the first time—and now this is just a blip on a cloudy day for CAISO’s dispatch of power to compensate for intermittency of solar power…if such fluctuations can be predicted as reliably as a solar eclipse can be predicted.

The ability to integrate renewables, and, going forward, to integrate electric vehicles of all kinds, both require tremendous improvements in the robustness and real-time capabilities of industrial data fabrics in support of advanced analytics, and in the digitation of the electric grid.

New Challenges vs. Our Ability to Address Them

At the conference, key challenges came to the forefront, along with great successes in meeting some of industries’ prior big challenges as exemplified by the integration of solar in California mentioned above.

Many of us have said it is good, in the energy transition, to take an “all of the above” approach, one where we support many different tracks on the path to resilient, economical, and reliable energy supplies. But the challenges that came to the forefront during this conference led me to look at the “all of the above” approach with some newfound skepticism. There are several reasons.

First, the use of certain clean energy solutions needs to be seen in a realistic context where different options are viewed and compared in practical ways. Some technologies are less viable than some people suggest, when it comes to their capacity to economically be reliable elements for new energy supply to meet all the new electric load coming on board. Second, it is not a question about such sources having sufficient technical feasibility. Instead, it is about their only being able to contribute a small portion of the needed capacity.

Two such examples are the use of farmland to create fuels off of the crops grown on the farmland, and second, the creation of biodiesel from reclaimed cooking oil:

• The energy efficiency of PV solar panels is about ten times higher than the rate at which plants convert sunlight to energy, and the use of farmland for “energy crops” instead of using it for crops in our food supply, should generally be shunned, since such high-cost energy sources also elevate food prices. And while making biodiesel from waste cooking oil is a better use of the waste cooking oil, than disposal of the oil in landfills, it is not going to be an important element of our energy mix.
• While my colleagues at the conference laughed when I pointed it out, they agreed with my suggestion that we would all have to increase our consumption of fried foods ten-fold to a hundred-fold, for biodiesel from cooking oil to be any sort of “pillar” of our clean energy supply. It should not be touted as such, but instead should be viewed, simply, as a better way, from a circular economy point of view, of “disposing” of waste cooking oil.

Data Centers, Hydrogen, and Transportation Electrification

Three big challenges discussed at the conference included the challenges associated with clean hydrogen, the promise of transportation electrification, and finally the unprecedented growth in electric power demand and T&D build-out associated with the massive new data centers quickly being added to our infrastructure.

The benefits of clean hydrogen, if it could be economically produced, are enormous. Hydrogen pilot projects of several major utilities were discussed, including LADWP and National Grid, along with interesting contributions from Southern Company regarding their positioning around hydrogen going forward.

The various technologies for, and challenges of, clean hydrogen were delved into in detail, including the vulnerability of welds in most of our natural gas utility pipe infrastructure to become embrittled if the amount of natural gas put in the mix exceeds 10%. The depth of industrial dependencies on hydrogen currently was highlighted, as was how small the current production levels were for clean hydrogen vs. the volumes of hydrogen in existing use cases overall.

Electrification of transportation will increase the economic and environmental benefits of clean electrification enormously, but will also require a massive increase in the ability of our infrastructure to utilize real-time data and control systems to run optimally and to create new markets for all participants (including drivers of passenger EVs and operators of EV freight trucking fleets). The complexity includes optimizing battery charging and routing of vehicles with widely varying needs.

A fascinating example of the complexities involved was around the benefits of regenerative braking systems on EV freight trucks, since such brakes not only help to charge the battery, but also help to extend the useful life of the friction brakes. If a truck is going to leave its charging station at a higher elevation than its first destination, its battery should not be charged to 100%, since it will exert undue wear on its brakes while going downhill and will lose an opportunity for some “free” charge off the regenerative brakes. A limiting case example given involved an actual use case of a massive EV truck at a mine at the top of a hill, which charged itself with regenerative braking while going downhill with a full lode of ore, and which was then able to go, empty, back up the hill “for free” in effect, not needing to have charging infrastructure installed for this use case.

The ability to build new electric infrastructure to charge large truck fleets is key, and is daunting, given timing, costs, and real estate issues involved. Many truck depots are in dense industrial areas where real estate to add new electric substation capacity may not be readily available. Many depots lease their space from landlords who have no interest in building out EV charging infrastructure. Seaports came to the forefront as a growing area where such infrastructure can more often be readily built, and where trucks are often going anyway to pick up cargo. Pilot programs for fast charging capabilities, and even for inductively charging with no physical connection between the charger and the truck, were discussed.

Finally, the other big challenge involved how new data centers are being built rapidly, and are enormous, with some coming in at 1,000 MW per data center (i.e., 10% of NYC’s total electric power demand, per data center!). The plans to build or upgrade needed substations and T&D infrastructure often go on fast tracks with these data center projects, based on the enormous demand for the compute power of these data centers, and the deep pockets of the investors involved. They typically, contractually, are expected to run at near 100% of their rated capacity, 24/7, so there is little or no option for most of them participating in peak demand reduction programs to help keep the grid up and running more reliably and economically—the exception is a smaller subset of data centers that are not dedicated to large language models or other AI-driven use cases, but instead have more flexibility to shift load if it is of economic benefit (e.g. this group includes Bitcoin mining data centers).

Industry’s Challenges are also its Opportunities

The three challenges mentioned above are interconnected and can be recast as opportunities.

While new installations of T&D switchgear can enable the faster control capabilities that are needed for a more dynamic grid to take on massive increases in electric vehicles, the existing infrastructure, including capacitor banks, involves electro-mechanical relays and control systems that cannot respond fast enough for optimal grid operations. Ironically, the tremendous growth spurts in demand for electricity to support new data centers for AI and cloud-based industrial data fabrics, functionally, will be enablers for the more dynamic grid, since the data center’s support for AI and data fabrics is precisely what is required to run the grid in the new ways it will need to be run—so utilities and their industrial, commercial, and residential end users will be among the key users of the new data center’s offerings.

Recommendations

Words matter, and the word “sustainability” has its place, but “risk reduction” is the better term here, in the context of recommendations for how to address our challenges better. And the risk of not stepping up is too great here, vs. the risk of stepping up, to meet the challenges, if they are seen clearly.

Risk reduction should be a cross-industry cultural mandate, and should ideally become embedded in a new generation of better regulatory and market-building structures. It became clear to me during the conference that we need as stern and strict a set of regulations being imposed on the local, state, and federal policymakers, as on industrial and public interest stakeholders. To meet 2050 goals requires a shared vision.

Enter full lifecycle costing capabilities, of a next-gen character. They should build upon, but go far beyond, what we used to call good old “engineering economics,” and should go hand in hand with build-out of big models to support better decision- and policy- making. In the context of energy transition and industrial sustainability, ARC recently articulated related trends with regard to a System of Systems (SOS) approach.

The best risk reductions are those that are entered into in a collaborative way, per the utility industry’s mutual aid assistance example. The business model upon which mutual assistance was built has been a great success. But it evolved in response to storm situations of the frequency and scale of the past, and, like other infrastructure and severe weather challenges, requires new refinements and improvements.

As was mentioned during the conference, the increasing frequency of line crews being called away to help in restoration efforts is starting to have negative impacts on the maintenance of the home utility’s T&D systems. A start related statistic is that there have been as many Category 4 and 5 hurricanes in the US in the last eight years, as in the prior 56 years—an 8-fold increase. Smarter T&D infrastructure, and better hardened T&D infrastructure, represents a shared resource of the industrial and commercial and residential end users who will be enabled to participate in new value areas of economic benefit to all participants.

If, and when, the new smarter T&D infrastructure gets built out, with the needed more real-time capabilities, there is still a regulatory disconnect that must be addressed, one for which, hopefully, the above-mentioned lifecycle costing, and big models, will be deployed. The utility example here has its counterpart in other industries: T&D infrastructure, and underlying baseload power generation infrastructure, has a 2x to 10x longer useful life, than the assets it is being connected to. Below, to provide context for the above recommendations, some examples are shown.

• If a truck depot electrifies, can the investors or regulators who fund and approve the T&D build-out be guaranteed that the electric capacity to charge the truck fleet will be needed after the (relatively much shorter) useful lives of those electric trucks have passed? What if hydrogen replaces electricity as the best fuel source for the freight trucks of the future?
• What if a regulator approves a microgrid at a location, whose economics were a function of the need for power at the end of a certain line, and the economics are ruined by some non-regulated build-out of power generating capacity from some competitor—what is the responsibility of the regulator to the microgrid owner who got permission to build the first facility?
• Similarly, what if some of new data centers being built this year are replaced at other locations by other data centers in a decade or two, at a point in time when the new T&D and generation infrastructure that was built to serve those data centers is still “young” and largely undepreciated?