Planning for Power and Long-Lead Equipment in Data Center Development | Provident Data Centers

Duration: 4:02 546 transcript words EN summary

Video summary

Justin Law, executive director of power and utility strategy at Provident Data Centers, emphasizes that approaching data center power development requires viewing the grid connection as part of a broader portfolio rather than a simple interconnection point. He argues that solutions implemented behind the meter must be comprehensive engineering feats with significant financial implications, not merely temporary fixes. To achieve this, developers should partner early with strong independent power producers to create holistic strategies that account for both technical complexity and long-term viability from the outset. The strategic location of a data center remains critical, particularly regarding access to existing infrastructure which serves as an essential bridge during development. Law highlights the importance of hybrid solutions that combine various power sources, noting that pairing reciprocating engines or BEZ units with AI workloads can provide robust and reliable bridging power. These setups are effective because they can eventually interconnect with the main grid, a feature that many end users still prioritize to ensure stability and scalability as their operations grow. Securing long-lead equipment has become increasingly challenging due to supply constraints faced by utilities and private substation builders. To mitigate these risks, Law advises developers to build transparent relationships with equipment suppliers by sharing their deployment pipelines and multi-year forecasts. This openness allows suppliers to adjust their production schedules accordingly, ensuring that necessary components are available when needed. Additionally, maintaining flexibility in equipment selection is crucial; while familiarity with certain brands is natural, the current pace of growth demands an open-minded approach to alternative options that offer technical feasibility, strong warranties, and reliable supply chains. Ultimately, de-risking hyperscale or inference development projects depends on a combination of strategic foresight, transparent communication with partners, and operational flexibility. By understanding their specific load profiles and designing solutions that anticipate future needs, developers can navigate the complexities of modern power infrastructure. The core message is that success in this sector relies on treating power planning as a sophisticated, long-term engineering challenge rather than a short-term utility task, requiring collaboration, adaptability, and a clear vision for the end state of the facility.

Read the full video transcript

Hey everybody. My name is Justin Law. I'm the executive director of power and utility strategy at Provident Data Centers. Yeah, that's a great question. So first, I think that what's important is to understand that when it comes to power, you think about it as a portfolio and not just a interconnection to the grid. [music] So when it comes to behind the meter power, I think it's really important to recognize that the the solution that you're doing is not just a band-aid, but it's an actually fully baked engineering solution that is very complex both from an engineer an engineering perspective and also a financial perspective. So I think it's important to be to begin with the end in mind and think about what that solution looks like. Work with a strong IPP independent power producer and generator company. Good developers are doing that these days and partnering early and also looking at the solution holistically. Location still matters. Hybrid solutions still are important and having access to the grid ultimately as a as a bridge is still important. So proximity to good infrastructure is a key component. And also finally back to the solution, it's important to understand what you're designing [music] for. It's not just bringing on generation. What's your load? What's your load [music] profile? And a lot of times reciprocating engines and and BEZ units paired with AI AI workloads is [music] a great way to have a strong robust highly reliable bridging power solution that is effective and also [music] can interconnect into the grid later, which a lot of end users are still seeking. Yeah, sure. So, long lead equipment is definitely one of those things that is getting more challenging either from a utility perspective and getting getting substation equipment either private or utility equipment or the vertical build themselves. And so, one important thing that you can you can do to position yourself is one build a relationship with the long lead equipment suppliers. Introduce yourself, be transparent with your portfolio, show them show them what your deployment pipeline looks like. And so, you know, showing them the forecast of what you're going to be needing in the next few years is as important as you getting it because they can >> [music] >> they can build for that into their their their forecasting schedule. So, that's that's one real great way is showing them the forecast. I would say the other thing that comes to my mind is being flexible with the equipment. Whether it's designing for a substation or doing, you know, directed chip liquid cooling, I think it's I think that most people are comfortable with the brands that they know just like people like the car the brand of cars that they drive. But, in a in a period of extreme growth that we have today, I think it's important to be flexible, look at other options, be open-minded [music] to other other solutions, other brands as long as it's technically feasible and you have you know, good warranties behind them, you have a strong supply chain, good good support. I think that's another great way to de-risk long lead equipment on any aspect of a hyperscale or inference development. Transparency.