Cross-zonal capacity (CZC) is a scarce resource. Every megawatt of transmission capacity reserved for balancing reserves is a megawatt that cannot be used by the day-ahead or intraday energy markets. To manage this, Europe has historically used a patchwork of different rules of the road.

In the first two parts of this series, we explored the highways of European energy balancing - the PICASSO and MARI platforms -  and the regional traffic managers like ALPACA, BBCM, and the Nordic Balancing Model (NBM) that coordinate reserves across borders. But even with the best highway and the most experienced managers, there is a fundamental question that remains: Who gets to use the fast lane?

In this final installment, we compare these methods and explain why the COBRA project’s harmonized approach is the key to a truly efficient European grid.

What are the different ways capacity is allocated for energy and balancing markets?

When cross-border trading first began, the methods for splitting capacity were relatively blunt. While these are rarely used today within major regional platforms, they set the stage for the complexity we face now.

Available Transmission Capacity (ATC) / Residual Method
Here, the wholesale energy markets get priority. The balancing market only gets whatever space on the road is left over after commercial trades have cleared. Balancing is the last line of defense. If the road is full, a TSO might find they have purchased reserves in a neighboring country that they physically cannot transport during a frequency emergency.

Fixed Reservation
This approach sets a permanent percentage of the highway (e.g., 10%) for balancing. It is the grid equivalent of an empty bus lane while the rest of the highway sits in a traffic jam. It wastes valuable capacity that could be used to lower wholesale energy prices.

This is why TSOs moved toward regional cooperation, with more sophisticated probabilistic and market-based methods.

The Probabilistic Method (ALPACA)
The ALPACA initiative (covering Austria, Germany, and Czechia) uses different strategies depending on the specific border:

On the AT-DE border, the TSOs (APG and the German TSOs) perform a monthly cost-benefit analysis that compares the day-ahead and aFRR market values of CZC, resulting in a market-based fixed allocation of up to 80 MW per delivery day. On the CZ-DE and AT-CZ borders, however, capacity is not explicitly reserved in advance. Instead, it uses a probabilistic method to forecast how much transmission capacity will actually be available in real-time without pre-allocating from the day-ahead market.

The Market-Based Method (NBM & BBCM)
Looking north, the Nordic and Baltic markets use market-based approaches under Article 41 of the EB Regulation (Electricity Balancing Regulation EU 2017/2195), which compares the value of capacity between the energy and balancing markets. Both markets ask: "Where is this capacity worth more right now: for energy trading or for securing reserves?" However, they answer the question in very different ways.

The Baltic Balancing Capacity Market (BBCM)
Its algorithm assesses CZC allocation based on the value of day-ahead trading versus the operational value of transferring reserved balancing capacity. The algorithm selects the allocation that yields the higher combined welfare across both markets. To manage this, it uses a default cap of 50% of the net transfer capacity.

The Nordic Balancing Model (NBM)
The NBM uses a sequential approach, where the balancing capacity market is cleared first. The model forecasts the value of CZC using price differences between countries based on a similar past reference day. To limit the impact on intraday and day-ahead trading, the reserved share of CZC is typically capped at around 10% of all available CZC.

What is wrong with current capacity allocation rules?

The primary reason we need a change is a structural conflict in how time is handled. Currently, capacity markets buy reserves and reserve transmission space before the day-ahead energy market even opens.

If the capacity platform assumes a certain amount of road space is available, but the real-time energy traffic is much heavier than expected, a gap emerges. The TSO is left holding a contract for a reserve they can't actually move across the border. Conversely, we might hold space for reserves that are never activated, wasting capacity that could have lowered prices for consumers.

How does harmonized market-based allocation through the COBRA project improve capacity allocation?

The COBRA Project (Common Optimization of Balancing Reserves and CZC Allocation), driven by ENTSO-E, addresses this gap at its root. It is important to note that COBRA will not create a new platform; rather, it is a harmonized algorithm - the Cross-Zonal Capacity Allocation Optimization Function (CZCAOF) - that can be used by all existing and future EU platforms that exchange balancing capacity and share reserves.

COBRA uses a Welfare-Maximizing Split to decide exactly how much CZC to allocate. It takes the actual firm bids from Balancing Service Providers (BSPs) and compares them against a high-fidelity forecast of the Day-Ahead Market (DAM).

The algorithm maximizes the total socio-economic welfare (W):

W = Economic Surplus (Balancing) + Economic Surplus (Day-Ahead Market)

If the math shows that reserving 50 MW for aFRR in a neighboring zone will save the system €10,000, but using that same 50 MW in the energy market will save €15,000, the algorithm will allocate the capacity to the energy market. 

COBRA moves us from a system of guessing and patching to a system of joint optimization. Its advantages are clear:

Consistency at Shared Borders 
Using three different methods creates the potential for inconsistency. As more regional balancing capacity platforms emerge across Europe, the same interconnector may eventually fall under the scope of two separate platforms. Without a shared algorithm, each platform would make independent CZC assumptions, potentially producing incoherent price signals across the grid. COBRA resolves this by ensuring every platform - regardless of region - runs the same CZCAOF blueprint. 

Expert Grid Modeling 
The algorithm is designed to handle both Net Transmission Capacity (NTC) and Flow-Based grid models. In flow-based regions, it uses Power Transfer Distribution Factors (PTDFs) to model how power physically moves across the grid. Crucially, it only takes into account burdening flows - a worst-case scenario approach that ensures the grid remains physically safe even if every single reserved megawatt is called into action at the exact same moment.

Accuracy through Forecast Validation 
COBRA includes a feedback loop. Regional Coordination Centres (RCCs) validate the forecasts after the market clears to determine the forecast error. This data is then used to refine the algorithm’s accuracy for future auctions.

Safety Valuations and Relaxation 
To protect the energy market, COBRA defaults to a 10% limit on CZC for balancing. However, this is not a hard wall. If the algorithm determines that a TSO’s safety demand cannot be satisfied within that limit, it can automatically relax the constraint to an extended limit to ensure the grid stays stable.

In extreme conditions, the algorithm might face infeasibility - a situation where grid limits and TSO demands simply cannot all be met at once. In these moments, the CZCAOF follows a strict mitigation hierarchy:

1. Increasing the CZC limits beyond the 10% default.
2. Using high-quality aFRR to fill a gap in the mFRR market if it is more cost-effective.
3. As a final resort, the algorithm will curtail TSO demand, signaling that the market cannot provide the required security.

Even when the market is stable, welfare ties can occur where two solutions are economically equal. To prevent randomness, COBRA uses deterministic rules: bids can be accepted pro-rata (split by volume) or based on their timestamp (the earliest bid gets priority).

What is the difference between sharing and exchanging balancing reserves?

The move to a harmonized algorithm also unlocks the full potential of ‘Sharing of Reserves’. While Exchange simply means buying a cheaper product in another country, Sharing recognizes that not every country will have a power plant fail at the same time.

COBRA allows TSOs to buy fewer reserves by coordinating that math. In the Baltic region, this logic helped deliver €470 million in annual savings, though that figure reflects a total system overhaul as the region left the isolated BRELL era. It was a leap from zero to one.

For the rest of the continent, the goal is precision. An @ACER study estimates that harmonizing this specific market-based approach will deliver €159 million in annual welfare gains. This is the final layer of polish on a system that is already running at high performance.

A unified map for Europe

The roadmap for the next few years is getting crowded. Between the massive load growth from data centers and the erratic nature of high-renewables generation, the competition for cross-border capacity is reaching a breaking point. TSOs are also staring down the 70% capacity floor requirement, a legal mandate that leaves little room for inefficient allocation.

The grid needs a single source of truth, which COBRA is set to provide. It handles the tension of the 70% goal by ensuring capacity moves where it creates the most welfare. This market-based approach is estimated to deliver €159 million in annual welfare gains. It is also the essential bridge toward full co-optimization, which could eventually unlock over €1.3 billion in system value. 

The data-driven, harmonized approach of COBRA will ensure that our continental power highway is managed with the precision required to keep the lights on while keeping energy affordable for all of Europe.

Navitasoft is a key partner in the digital transformation of European energy. Our expertise in balancing capacity platforms like the BBCM ensures that our clients are ready for the harmonized, algorithmic future of the grid.