Energy Market Management

Forecast bidding is no longer the obvious default. You run a wind farm. Every day before noon, you decide how much power to sell in the day-ahead electricity market. You can bid your best production forecast. Or you can deliberately deviate and bet on the spread between day-ahead and balancing prices. For years, that bet was barely worth making. Europe's old two-price balancing schemes penalized imbalances by direction, which limited arbitrage. But the market has changed. With the move to single-price balancing and new balancing price formation, now shaped by both mFRR and aFRR activation markets, balancing prices have become far more volatile. The spread between day-ahead and balancing prices is now wider, less predictable, and potentially profitable. But also risky. In our new manuscript, we propose a data-driven bidding framework that answers three questions: → When to engage in arbitrage. A probabilistic classifier decides whether the predicted spread is confident enough to act on. Otherwise, you bid your forecast. → Which direction. When engaging in arbitrage, the classifier also tells you whether to go long or short. → How much to arbitrage. A linear policy, learned through contextual optimization, maps contextual features to the arbitrage size, while a CVaR constraint keeps tail risk in check. Using real Danish DK1 and German DE/LU data, the strategy improves mean profit over forecast bidding, by around 7% for a hybrid wind-electrolyzer plant in DK1, for instance. The hybrid plant earns more from arbitrage than wind alone, because the electrolyzer absorbs part of the mismatch internally. That creates more room to trade without sending every deviation to the balancing market. But there is a limit. The gains are strongest when recent conditions still resemble the training period. But the environment is non-stationary: when it shifts, past data becomes a misleading guide and the edge shrinks. This does not explain every bad window, but it points to a real weakness of data-driven arbitrage, one you only see after the fact. Joint work with Yannick H. and Farzaneh Pourahmadi. That is where the project goes next. During an upcoming PhD research stay with David Wozabal, Yannick will build distributional robustness into the framework, so the bidding policy can hedge directly against drift. Preprint: https://lnkd.in/dX2J-ZGt DTU Wind and Energy Systems

Maximising BESS revenue requires fully automated aFRR EAM trading. For a battery energy storage system (BESS), marginal cost isn't driven by hardware or fuel costs; it's driven by the aggregated volatility across all accessible markets. Alaa Farhat Noureddine and I have been looking into our trade data and visualised our realised activation prices for one of our 2-hour BESS assets (normalised to 1 MW/2 MWh). Here's what the data shows: 🔋 The real revenue driver is energy, not capacity The primary revenue driver of aFRR is not the capacity market. It is discharging energy into the Energy Activation Market (EAM) as up-regulation activations, which pay significantly above the spot price. Discharging via EAM up and recharging at spot yields a spread of 272 EUR/MWh, nearly double the day-ahead average spread of 137 EUR/MWh.   🔋 Recharging: where the real optimisation happens   Recharging via EAM down-regulation reduces buy costs, increasing the spread from 272 to 310 EUR/MWh. Down-regulation prices are often negative, but recharging at any price below spot improves the financial result. In the period, we achieved a 55% reduction in recharging costs compared to spot.   86–126% of discharged energy from EAM up-activations is recharged through the down-regulation market, at a cost of just 12–16% of the activation revenue. The remaining energy is purchased on the intraday continuous (IDC) market. 🔋 Why this require full automation The EAM is a marginal-priced market, so optimal bidding means offering your true marginal cost to the up market: the expected cost of recharging after the activation, based on current and forecasted prices in the IDC and EAM down. This requires continuously updating asset-level bids based on real-time market conditions, something no human trader can do manually. It lets you offer energy to the system as cheaply as possible while ensuring every realised activation is financially beneficial. A fully automated, real-time trading system captures value that manual strategies leave on the table. If your asset isn't capturing this value, reach out. We can quickly onboard you onto our fully automated real-time trading and optimisation platform. Hybrid Greentech - Energy Storage Intelligence

VPPs Are All the Rage – But They’re Not Just for Households! ⚡ Virtual Power Plants (#VPP) are once again a hot topic—and for good reason! The focus often falls on aggregating smaller players, like households or small producers, into a unified power source. However, the VPP model is just as relevant for large-scale producers managing a portfolio of Power Purchase Agreements (#PPA) from renewable assets like wind, solar, and storage. By treating renewable assets as an integrated portfolio, substantial value can be unlocked. Additionally, centralized portfolio management helps protect revenue against the volatile effects of renewable-dominated markets Turning Your PPA Bundle into a VPP  Managing a portfolio of PPAs from wind, solar, and storage assets mirrors the process of a “small” VPP. Through technology, these assets can be interconnected which then allows for the optimization across various energy markets, from ancillary services to bilateral PPAs. This portfolio approach maximizes the efficiency of diverse assets through centralized control, just like a VPP. How to Transform Your PPA Portfolio into a VPP 1. Digitally Connect Your Assets Gain the ability to operate your units as a single entity by connecting them through infrastructure and software, which are readily available and proven effective. 2. Build a Dedicated Commercial Team Start with a revenue management strategy that covers the full spectrum of PPA durations—from long-term contracts to day-ahead markets and ancillary services. This specialized team should structure, price, and execute PPA, hedging, and trading strategies. Most of the execution work can be outsourced as well, but oversight and control over partners remain essential 3. Enhance Data and Analytics Implement systems that offer deep insights into revenue streams, risk profiles, and market changes' impacts. Robust data and analytics are essential to managing a dynamic portfolio. The Benefits of Operating a Large-Scale VPP  A large-scale renewable portfolio managed as a VPP—even one based on long-term PPAs—can drive meaningful savings through reduced Route-to-Market and balancing costs while generating additional revenues. These gains arise from the flexibility to optimize production across all available energy markets. Most importantly, this approach allows producers to participate in future markets and innovative business models, such as offering fixed green shapes (see my recent post on 7/11 PPAs), selling power to smaller but higher-yielding industrial off-takers, and mitigating the impact of negative prices. Transforming a PPA portfolio into a VPP will require a dedicated effort, a clear commitment from top management, and an understanding that the journey will be a longer-term one. Embracing this approach positions renewable portfolios to thrive in the evolving energy landscape while unlocking new potential for sustained growth.

Effects of elastic demand in PICASSO/MARI are worth watching. Several EU TSOs are already applying elastic demand to their aFRR or mFRR bids, with more to join. The outcomes can potentially be felt way beyond their own LFC areas. Why? PICASSO and MARI moved balancing energy activation markets from local merit-order lists into common European optimisation based on common merit order lists. BUT: If TSO demand is treated as fully inelastic (as it traditionally has been), the logic is to activate the requested balancing volume “at any price”. In scarcity situations or basically when only very expensive bids are available this can quickly can produce extreme activation prices even for volumes that go beyond what is strictly needed for system security purposes. ACER 2024 Decision concerning the aFRR/PICASSO implementation framework under EBGL Art. 21 corrected that (MARI IF already had in place earlier). In practical terms, elastic demand turns PART of the TSO’s balancing energy demand from a vertical curve into a price-sensitive curve. For all TSOs applying it, elastic demand is a kind of safety valve but the calibration of the curve is rather different. For BRPs, it can reduce exposure to extreme imbalance price outcomes although it also makes the link between system imbalance, balancing activation and imbalance price formation more complex. For BSPs, very high-priced bids may no longer be activated just because the TSO has submitted demand. I.e. activation probability for those bids goes down & the risk model changes. The farther PICASSO/MARI expansion continues, the more important the shape of TSO demand curves becomes for BSP revenues, BRP exposure and investment cases for flexibility. If there are more TSOs applying elastic demand that I'm unaware of, please do let me know! #balancing #balancingenergy #mFRR #aFRR #PICASSO #MARI #BRP #BSP #flexibility

This week's the National Electricity Market milestone got a lot of coverage. The dispatch optimisation question got almost none. The NEM, Australia's (National Electricity Market), just approved full operations for its first 8-hour battery, RWE's Limondale BESS in southern NSW. 50MW, 400MWh, 144 Tesla Megapacks. It's a significant milestone. Australia is now the world's third-largest utility-scale battery market. Renewables hit 43% of Australia's electricity in 2025. The hardware investment is real. The operational challenge is what comes next. A BESS (battery energy storage system), generates revenue through charge and discharge decisions made across multiple markets simultaneously. Wholesale spot prices, frequency control ancillary services, capacity markets, arbitrage windows. Every hour, those conditions change. Every hour, the optimal strategy changes with them. Statkraft Croatia used Uprise d.o.o. and Gurobi Optimization to model exactly this. The problem: plan a 20-25 year BESS asset across volatile prices, ancillary service rules, battery degradation, and multiple market strategies. The model type: MILP (Mixed Integer Linear Programming) hourly multimarket optimisation. The result: 12% to 25% increase in forecasted revenue, plus a detailed hourly trading strategy to support investment decisions. The solver ran daily. In energy markets, calculations often happen multiple times a day. Speed isn't a nice-to-have. It's a commercial requirement. As Australia's battery fleet grows, the question shifts from whether to build, to how well you can dispatch what you've built. What does your team's BESS dispatch strategy look like today, and is it keeping pace with NEM price volatility? 👇 Comment "BATTERY" and I'll send you the full PDF Statkraft battery optimisation breakdown 📚 #BatteryStorage #EnergyOptimization #MathematicalOptimization #NEM #bess #NationalElectricityMarket #ElectricityMarket #Electricity #BatteryOptimisation #MathematicalOptimisation

The Nordic battery storage market is experiencing a massive boom, with over 4GW of new capacity entering the pipeline in the last year alone, according to Aurora Energy Research. Sweden, particularly in the SE3 and SE4 price zones, is leading this charge. While this signals a huge step forward for our energy transition, it presents a critical challenge for asset owners. This unprecedented growth is creating market saturation, causing ancillary service prices—the traditional bread and butter for BESS revenues—to decline. In Finland, for example, we've seen a recent drop in revenues for all BESS durations. Relying on a single revenue stream like Frequency Containment Reserve (FCR) is no longer a viable strategy for guaranteeing a strong IRR. The era of passive asset management is over. To thrive in this evolving landscape, profitability now hinges on sophisticated, multi-market optimisation. Asset owners must look beyond ancillary services and strategically stack revenues from different markets, including day-ahead trading, emerging flexibility services like mFRR and aFRR, and capturing high-volatility spreads. It's no longer just about the size of the battery; it's about the intelligence of the strategy behind it. The key to maximising returns lies in dynamic optimisation that can navigate market saturation and unlock hidden value. For asset owners in the Nordics, how are you adapting your strategy to protect your returns in this rapidly maturing market? #GreenVoltis #BESS #EnergyMarkets #AssetOptimization #RenewableEnergy #Nordics