Obstacles in Grid Capacity Data Management

“Despite having some 10 million smart meters, the electric utilities named in the complaint are unwilling or unable to provide customer power usage data to aggregators such as Voltus, known as Curtailment Service Providers (CSPs). For each of eight major utilities in the PJM region, the complaint documents its attempts at accessing smart meter data. The process is described as “byzantine,” outright unavailable, or so onerous that enrolling thousands of homes in a virtual power plant would be impossible. These allegations are consistent with Mission:data’s Green Button Explorer, which found that not a single utility in PJM’s footprint provides the customer data necessary for behind-the-meter resources to deliver energy and capacity at lower prices. Today’s filing is the second attempt ever to bring data access barriers to FERC’s attention. In 2023, CPower filed a complaint against PJM regarding similar topics. That complaint was denied without prejudice in September 2024 for lack of evidence. Today’s filing seeks to overcome the evidentiary hurdle by meticulously documenting widespread practices of data blocking.”

Europe’s grid runs on PDFs and emails while solar waits in line. It’s time for digital change. 𝗟𝗼𝗼𝗸 𝗮𝘁 𝘁𝗵𝗲 𝗻𝘂𝗺𝗯𝗲𝗿𝘀: – Over 1,700 GW stuck in grid queues – That’s 3× what Europe needs by 2030 – Most of it blocked by process, not physics I’ve seen this story play out for more than a decade. Projects don’t fail because copper is missing. Projects fail because structure is missing. What’s the real bottleneck in 2025? It’s not cables. It’s the process grid. Here are the five changes Europe needs: 𝗗𝗶𝗴𝗶𝘁𝗮𝗹𝗶𝘀𝗮𝘁𝗶𝗼𝗻 Every DSO and TSO runs its own system. Some use digital portals. Some use PDFs and email attachments. Some use Excel sheets. In real projects, that means: – No real-time status – No unified data model – No automatic checks – No transparency A single misplaced document can cost weeks. Spain and Germany now require digital portals. EU-wide? Still far from a shared system. A digital EU standard would remove friction. 𝗘𝗨-𝘄𝗶𝗱𝗲 𝗛𝗮𝗿𝗺𝗼𝗻𝗶𝘀𝗮𝘁𝗶𝗼𝗻 Europe has: – 40 TSOs – 900+ DSOs in EU-27 – 2,500+ DGOs All with different forms, processes, timelines. For developers in multiple countries, fragmentation is the true bottleneck. Scaling to 2030/2040 targets? Impossible without harmonisation. One form logic. One process architecture. Not 27 variations. 𝗧𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝗰𝘆 𝗼𝗳 𝗤𝘂𝗲𝘂𝗲 & 𝗣𝗿𝗶𝗼𝗿𝗶𝘁𝗶𝘀𝗮𝘁𝗶𝗼𝗻 Most countries do not show: – Which projects are in the queue – How long each step takes – How capacity is allocated – How priorities are set Ghost projects block capacity for years. In the UK, up to 70% of queue entries are non-viable. Planning becomes guesswork. Guesswork is expensive. A simple EU Queue Dashboard would change the game. 𝗣𝗿𝗲𝗱𝗶𝗰𝘁𝗮𝗯𝗶𝗹𝗶𝘁𝘆 Speed is not the issue. Predictability is. If step X takes 8 weeks, say it clearly. Developers can manage long timelines. They cannot manage uncertainty. Uncertainty raises CAPEX, delays financial close, blocks risk evaluation. A predictable, EU-wide process is worth more than isolated acceleration. 𝗖𝗹𝗲𝗮𝗿 𝗥𝗼𝗹𝗲𝘀 (𝗗𝗦𝗢 / 𝗧𝗦𝗢 / 𝗡𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝘂𝘁𝗵𝗼𝗿𝗶𝘁𝗶𝗲𝘀) Responsibilities are defined on paper. But every country implements them differently. Typical results: – Delays – Repeated studies – Unclear decision paths The new EU Grid Package tries to address this. But implementation will take years. Until then, Europe runs a system designed for another era. We need a harmonised responsibility model, defined by a digital platform. 𝗧𝗵𝗲 𝗯𝗼𝘁𝘁𝗼𝗺 𝗹𝗶𝗻𝗲 If Europe wants to unlock the 1,700-GW pipeline, the priority is not “more cables”. The priority is: – Digital baseline – Harmonised rules – Full transparency – Predictable timelines – Clear responsibilities Where do you see the biggest lever for process change? What would be your first mandatory feature in a digital, EU-wide grid connection portal-and why? #AndreasBach #GridReality

Grid #connections are slowing down Europe’s clean energy transition, but there are solutions. Last week, the European Commission published the #GridsPackage, which includes a Guidance on Efficient Grid Connections. Here are some crucial points the Guidance addresses: ⚠️ Root causes of connection backlogs: ⁉️ Grid planning & development: Building new #grid #capacity takes 4–10 years, while new connection projects are usually ready in 2-3 years. Delays are due to a combination of slow #permitting procedures, #supplychain issues for certain components, and limited labor capacity. ⁉️ Transparency gaps: Customers often lack knowledge of available capacity due to limited data, particularly at the low-voltage level. Additionally, only a few Member States adopt tariffs that push connection requests to less congested areas. ⁉️ Process inefficiencies: Including first-come, first-served queues, manual processing, and speculative applications. Only 11 Member States have fully digital procedures. ✅ Promising solutions ✔️Greater #transparency and #efficient use of the grid: High-quality #hostingcapacity #maps can help plan where to connect. The EU is developing “Capacitypedia,” a pan-European platform for capacity maps and data. #Dynamictariffs, locational charges, and auditing unused reservations help free up capacity. For example, our recent blog helps explain why grid capacity data is key to electrifying road freight: https://lnkd.in/dG8QzVxq ✔️ #Prioritisation frameworks: Moving beyond first-come, first-served is possible under EU law. Criteria can include grid-friendly uses, social infrastructure, and housing, among others. ✔️ Modernized connection procedures: #Digitalising the process can speed up approvals. AI can validate data and calculate capacity, while digital twins can simulate grids to support faster and more informed decisions. Milestone-based rules, deposits, and planning prerequisites reduce speculative applications. Some countries also batch requests to streamline queues. This guidance is a crucial component of the Grids Package: timely grid access can help minimize potential connection delays, which impact electrification across all economic sectors, including industry, transportation, and data infrastructure. Addressing those issues will be crucial for meeting climate neutrality by 2050. The International Council on Clean Transportation

Secure capacity for Data Centers is diminishing. Short-term measures to alleviate Europe’s grid capacity challenges mean new and existing data centres increasingly need to actively win and retain capacity. Data Center growth strategies will need to adapt to the evolving environment. The short-term measures - different approaches in different markets, ranging from testing non-firm (flexible) connection contracts to enforcing strict "claw-back" mechanisms for unused capacity - are not a perfect solution. Nevertheless, they are unlocking meaningful capacity right now, bridging the gap while new infrastructure is still years away. What's more their success is rapidly leading to them becoming mainstream. The implications for data centers can be significant: - New data center projects need to adopt flexible connections to secure timely access. In constrained areas this may be the only way in. - Existing data centers must monitor utilization closely, as unused capacity risks being reclaimed under emerging "use-it-or-lose-it" rules. The following map shows the extent to which data center operations are being impacted around Europe. In High Impact Markets both claw-back measures and flexible contracts options are in place, and flexible contracts are the only pathway for grid access. In Moderate Impact Markets both claw-back measures and flexible contracts exist. In Low Impact Markets some form of claw-back mechanisms are present. In Markets with Likely No Impact, there is no observable measures in place so far.   For more information feel free to reach out to our head of grid related services Manohar Bayyapu Reddy #DataCenter #GridCapacity #Europe #EnergyFlexibility

Malaysia's data centre boom reveals a fascinating puzzle: facilities are currently using less than half (47%) of their declared maximum electricity demand. This gap between projected and actual consumption is not just a statistic; it is a strategic risk, creating the potential for stranded assets and higher costs for everyone. How do we accurately forecast monumental resource needs ? Think of it like building a highway based on peak holiday traffic projections for every single day. The infrastructure must be built, but if the cars don't show up, the investment becomes inefficient for the builder and costly for the public. This is the "Phantom Capacity" dilemma. The core challenge is not about curbing ambition, but about sharpening its accuracy. Over-declaration, whether from speculative applications or overly cautious planning, can strain a nation's entire power grid stability and economic planning. The solution lies in moving from static declarations to dynamic, evidence-based forecasting. This requires a collaborative model between industry and regulators. For Operators: It is about leveraging deeper big data—from real-time server utilization to committed customer roll-out plans—to build more granular power models based on big data collected on electrical assets. For Governance: It involves creating a robust framework that validates demand based on tangible milestones and track records, much like how financial institutions assess creditworthiness. Developing an effective Maximum Demand Declaration management system is not about stifling growth, but about ensuring its sustainability. It is about building a digital infrastructure that is both powerful and prudent, ensuring that our nation's energy is allocated as efficiently as the data flowing through the servers themselves. #DataCentres #EnergyEfficiency #SustainableGrowth #BigData #EnergyMonitoring #DigitalInfrastructure #Malaysia #StrandedAssets #EnergyPolicy

Capacity vs. Energy for the AI Boom: Why the grid is sweating As this chart from ISO New England shows, there is a fundamental distinction between Capacity and Energy and if you don't understand it, you don't understand why the AI boom might break the power market. Everyone talks about how much "energy" data centers consume. They point to annual TWh (Terawatt-hour) figures and worry about the total volume. They’re looking at the wrong bucket. We aren't running out of energy; we are running out of Capacity. 𝗧𝗵𝗲 𝗦𝗽𝗿𝗶𝗻𝘁 𝘃𝘀. 𝗧𝗵𝗲 𝗠𝗮𝗿𝗮𝘁𝗵𝗼𝗻 Think of it like a highway: Energy is the total number of cars that travel the road in a year. Capacity is the number of lanes. It doesn’t matter if the road is empty 23 hours a day; if 10,000 cars try to hit the pavement at 5:00 PM and you only have two lanes, the system fails. AI is a fleet of heavy-duty trucks trying to merge into the HOV lane during rush hour, 24/7. 𝗪𝗵𝘆 𝘁𝗵𝗶𝘀 𝗶𝘀 𝗮 𝗖𝗮𝗽𝗮𝗰𝗶𝘁𝘆 𝗖𝗿𝘂𝗻𝗰𝗵 1- The "Always On" Burden: Forget the "flat" load of traditional data centers. AI is an industrial baseload monster driven by the unforgiving ROI of $40,000 H100s. When you're sitting on a $10B compute cluster, every second of idle time is a catastrophic capital loss. Training runs don't sleep; they cycle 24/7/365. This creates a "strobe light" effect on the grid: massive, bursty spikes of hundreds of megawatts as models "checkpoint," leaving ISOs with zero margin for maintenance or recovery. For the grid operator, this isn't just a load increase; it’s a high-frequency volatility nightmare. 2- The Dispatchability Gap: In many ISO's we are retiring or planning to retire "dispatchable" capacity (coal/gas) and replacing it with "intermittent" energy (wind/solar). While we have plenty of energy when the sun shines, our capacity to meet a sudden peak is shrinking. This is where batteries and demand response will come in. 𝗧𝗵𝗲 𝗕𝗼𝘁𝘁𝗼𝗺 𝗟𝗶𝗻𝗲 The grid is moving from a world of "how much fuel do we have?" to "how much instantaneous load can we handle?" Reliability isn't about the total GWh generated over a year; it’s about the MW available at the exact millisecond a transformer is about to blow. We are currently building a digital economy that assumes infinite "lanes" on a physical grid that is already narrow and crumbling. You can build a 1GW wind farm, but if the transmission lines into the RTO/ISO are throttled, that capacity is "stranded." We are seeing "Interconnection Queues" stretching into the 2030s because the physical hardware, transformers and high-voltage lines, cannot be willed into existence. Software moves fast, but grid evolution takes time. This is why many grids are opting for the Bring Your Own Capacity model. 📌 Are we fast-tracking a "Capacity Armageddon"? What do you think will end up as the main way to power AI data centers?