10 Economic variants methodology //

The economic variants – the High Economic Variant (HEV) and the Low Economic Variant (LEV) – are built from the NT dataset, prior to gap-filling. This sequencing is a deliberate methodological constraint: the variants are constructed at the ETM stage, using the same modelling environment in which TSOs developed the NT dataset, before the policy-compliance transformation that produces NT+ is applied. Therefore, the economic variants are not required to comply with EU climate and energy targets, as their role is to explore sensitivities to macroeconomic conditions rather than policy-consistent pathways. Their purpose is different: in line with ACER Framework Guidelines Article 37, HEV and LEV are designed as plausible stress tests of the NT energy system, exploring how different macroeconomic and socioeconomic conditions may influence energy demand, technology uptake, sectoral activity and marginal-cost signals relative to the NT baseline.

This distinction carries two important implications that the reader should keep in mind throughout this chapter. First, all scaling factors, baseline shares and annual reference values used in the variant construction methodology are computed relative to NT values – not NT+ values.

Second, the economic variants do not represent standalone policy scenarios, nor are they intended to sit within the policy-compliance envelope defined by the NT+. They complement the Central Scenario by providing a bounded sensitivity analysis around the NT baseline, supporting a more robust assessment of infrastructure needs under different economic conditions.

For TYNDP 2026, HEV and LEV are applied to the 2035 and 2040 target horizons. It is also important to acknowledge the scope of what these variants do and do not capture. Their trajectories remain relatively close to NT and are therefore less divergent than alternative scenarios used in past TYNDP cycles such as Distributed Energy and Global Ambition. They should not be interpreted as an exhaustive exploration of the full range of possible future developments.

Table 26 below summarises all parameters, their treatment in each variant, and whether they are exogenously set, model-determined, or fixed. It is placed here deliberately to give readers an immediate, concrete picture of what changes before the methodological narrative unfolds.

The economic variants are applied to the 2035 and 2040 horizons, which correspond to the mid-term and long-term planning horizons required for the TYNDP framework.

These time horizons are sufficiently distant for economic divergence to materially affect energy demand, technology deployment, and infrastructure requirements, while still being relevant for network planning decisions.

By contrast, the 2030 horizon is treated as a short-term milestone and remains closer to the implementation logic of current policies and near-term planning assumptions. The 2050 horizon remains part of the overall scenario timeline, but for the economic variants exercise it is considered very long-term, with significant uncertainty. For this reason, the economic variant framework is primarily centred on 2035 and 2040, where the value of testing different economic conditions becomes more meaningful.

The high economic growth variant reflects a context in which economic performance exceeds the assumptions embedded in the central scenario (Chapter 9 of the Scenario Report). In practical terms, this variant is associated with higher GDP, stronger sectoral activity, greater purchasing capacity, and a higher willingness to spend and invest. It also assumes a stronger orientation towards innovation, including riskier or earlier-stage investments, and a greater capacity to take long-term decisions in support of sustainability-related investments.

By contrast, the low economic growth variant reflects a more constrained economic environment. In this case, GDP growth is weaker than in the central scenario. Sectoral activity is more moderate, spending capacity is reduced, and consumers and firms are more likely to prioritise affordability and short-term cost control.

Under these conditions, investment appetite may weaken, innovation may proceed more slowly, and decision-making may favour business-as-usual or lower-risk pathways.

These contrasting assumptions are not introduced to build optimistic or pessimistic narratives for their own sake. Rather, they are intended to capture the main economic mechanisms that may alter the pace and scale of energy transition drivers. A stronger economy can accelerate demand growth, industrial output, technology deployment, and investment in new assets. A weaker economy can delay replacement cycles, reduce activity levels, and shift choices towards lower-cost solutions. The variants therefore test not only volume effects, but also behavioural and structural responses that are relevant for infrastructure planning.

(i) Variants are anchored to the central scenario

The starting point for both economic variants is always the Central Scenario. This ensures methodological consistency across all scenario outcomes and preserves the role of National Trends as the main reference case for TYNDP 2026.

Anchoring the variants to the central scenario has two methodological benefits. First, it avoids the creation of disconnected scenario worlds based on entirely separate assumptions. Second, it makes the effect of economic deviations easier to interpret, since changes observed in the variants can be read an outcome of adjusted economic conditions rather than as the combined result of multiple unrelated methodological changes.

In this sense, the variants should be understood as stress tests of the central scenario, not as standalone products.

(ii) Variations remain limited and plausible

The purpose of the economic variants is to explore plausible economic deviations, not extreme macroeconomic futures. The deviations from the central scenario therefore remain within ranges that are considered credible and relevant for infrastructure planning.

This design choice ensures that the variants remain both informative and credible. Variants that are too close to the central scenario would add limited analytical value, while variants based on extreme assumptions could undermine the relevance of the results for planning decisions.

The methodology therefore balances analytical differentiation with realistic economic conditions. This approach also supports comparability and transparency across countries and sectors.

(iii) Targeted variation of key parameters

The economic variants are implemented through targeted changes in selected key parameters that transmit economic conditions into the energy system.

The objective is not to construct a completely new storyline, but to adjust parameters that have a clear relationship with economic activity and energy demand. These parameters may include, for example, activity-related drivers, technology uptake rates, investment-related assumptions, demand volumes in selected sectors, and behavioural parameters associated with affordability or spending willingness.

Focusing on a limited number of influential parameters ensures that the variants remain transparent and manageable while still capturing meaningful sensitivities in the modelling results. The full list of parameters selected for variation, together with those held fixed and the rationale, is documented in Annex IX.

(iv) Balanced contrasts across variants

Where appropriate, parameter changes should be applied in a balanced way across the two variants. In other words, when a relevant parameter is adjusted upward in the high variant, the methodology should assess whether a corresponding downward adjustment is justified in the low variant, and vice versa.

This principle supports transparent comparison between variants and facilitates interpretation of the modelling results.

However, balanced contrasts do not imply strict symmetry in all cases. In some sectors, economic responses may be non-linear or constrained by technological or policy factors. The changes in the parameters are subject to the Saturation curve (S-curve) methodology, which allows flexibility where needed to preserve realistic system behaviour.

(v) Saturation methodology for adjusting technology mixes in economic variants

As previously described, the economic variants are derived from the NT dataset by changing two elements: the level of sectoral activity and the technology mix. Sectoral activity describes the volume of activity in a sector (for example, total passenger kilometres driven by personal vehicles) and can be scaled directly. The technology mix is the disaggregation of that sectoral activity across technologies and / or energy carriers (for example, the shares of battery electric, hydrogen, and fossil fuels in road transport). Because each technology has its own efficiency, changing the mix implicitly alters aggregate system efficiency.

To keep adjustments realistic, the methodology embeds observed adoption and phaseout dynamics. Adoption typically resembles an S-curve: low and slow uptake in the early phase, rapid growth during diffusion, then slowing growth as saturation is approached. Phaseout shows resistance: as a technology declines, a residual “core” use remains that is hard to substitute, so further decreases become progressively limited.

Rather than assigning discrete stages by judgment, the stage and the feasible adjustment are determined directly from the current market share s (in percent) using two uniform functions that are applied consistently across all sectors and target years:

These functions eliminate the need for technology or sector specific parameterisation, ensuring transparency and avoiding unnecessary complexity. They also implicitly classify technologies by stage: low f_{\mathrm{up}}(s) indicates early or saturated conditions, high f_{\mathrm{up}}(s) indicates active diffusion; low of f_{\mathrm{down}}(s) indicates a residual core that should not be pushed down further, a high f_{\mathrm{down}}(s) indicates ample scope to reduce.

The adjustment procedure proceeds as follows. First, the economic variant narrative indicates which technologies should increase (for example, electric and hydrogen) and which must decrease to compensate (for example, diesel and gasoline), preserving the constraint that the sum of shares remains 100 %. Using the NT baseline shares, each technology’s feasible change is computed by multiplying the technology mix scaler with f_{\mathrm{up}}(s) for each technology slated to increase and f_{\mathrm{down}}(s) for those slated to decrease. The total feasible increase and the total feasible decrease are then reconciled. If one side’s potential exceeds the other’s, the larger side is proportionally scaled to the smaller side (the bottleneck), preserving balance.

Adjustments are distributed proportionally to each technology’s computed capacity, applied to the shares, and the mix is normalised to ensure it sums to 100 % without exceeding practical bounds implied by the functions. A numerical illustration of this procedure applied to a specific sector and country is provided in Annex IX, alongside the complete parameter adjustment outputs for the 2035 and 2040 horizons.

This approach yields smooth, plausible trajectories: early-stage technologies can grow, but not unrealistically fast; mid-stage technologies grow most strongly; saturated technologies slow down; and declining technologies reduce progressively without being forced below practical minima. By using the same functional forms across sectors and years, the methodology remains consistent and reproducible while capturing the essential saturation and phaseout behaviours observed in real markets.

(vi) Scaling electricity and hydrogen hourly demand

As previously mentioned, the NT dataset’s hourly electricity demand profiles were built using DFT starting from annual demand values collected from ETM and from the second data collection. Hourly electricity demand input values were divided in three main categories in PLEXOS®:

The NT dataset’s hourly load profiles (divided in the three main categories listed above) were then rescaled using “scaling factors” to derive the new PLEXOS® inputs for the high and low economic variants.

The scaling factors for each demand category were defined as the ratio of economic variants’ annual demand values over its corresponding annual value from the NT dataset.

For example, electricity demand for passenger electric vehicles in Germany in 2040 on ETM was estimated at 110.58 TWh, 125.07 TWh and 96.26 TWh in the NT dataset, High variant and Low variant respectively. Therefore, the 2040 factors that were used to rescale EVs hourly demand values in Germany were equal to 1.1311 (or +13.11 %) for the high variant and 0.8705 (or –12.95 %) for the low variant.

An alternative methodology was implemented for the members that chose to use ERAA demand values rather than ETM outputs (e. g., Bulgaria and Cyprus). Total electricity load in PLEXOS® was not split into different demand categories in PLEXOS® for those countries. The scaling factors for those states were defined as the average scaling factors of their neighbouring countries weighted by their total load.

The methodology described in this chapter was also used to rescale hydrogen hourly demand profiles, except that a single scaling factor per scenario was used. (In PLEXOS®, hydrogen demand input values are not split into multiple categories).

The same hydrogen demand scaling factors for Romania were also used to define the new hourly inputs for Bulgaria.

Hydrogen demand in the United Kingdom was increased /decreased based on the average scaling factors of France and Germany.

In the design of the economic variants, the primary objective is to stress test the central scenario’s supply capacities under different levels of demand driven by varying macro and socio economic conditions (i. e. different levels of economic activity and growth). The supply side infrastructure (installed assets and capacities) is therefore held fixed across variants, and the variants explore how this fixed system performs when demand and marginal costs conditions deviate from the central case.

From the supply perspective, several items could in principle be affected by economic variants: commodity costs for fossil fuels, the CO₂ price, energy supply capacities, flexibility providing capacities, cost of capital (WACC), technology costs, CCU / S volumes, and energy imports. In practice, for this cycle, only cost parameters are varied. Installed capacities (generation, conversion, flexibility, import infrastructure) are not expanded or reduced in any of the economic variants, and their dispatch remains fully determined endogenously by the model in all cases.

WACC and technology costs are not adjusted either, as no endogenous capacity expansion is modelled in this cycle and changes in investment costs would therefore not affect the asset base. Likewise, CCU / S volumes are not set ex-ante but continue to be purely model outputs, driven by the interplay of the fixed infrastructure, demand levels and marginal cost signals.

For energy imports, the physical import infrastructure and its technical limits remain unchanged but import flows respond endogenously to the different demand levels and marginal cost assumptions in each variant.

The main levers used to reflect different economic conditions are commodity and CO₂ prices. Their values are modified in line with the high level narrative of each variant: in a higher growth environment, stronger overall energy demand and emission allowances translates into higher prices; in a lower growth environment, weaker demand leads to lower prices. Concretely, a constant proportional adjustment is applied to both commodity and CO₂ prices: +10 % in the High Economic variant and −10 % in the Lower Economic variant relative to the central scenario.

Under this setup, the economic variants do not alter the structure or capacity of the supply system. Instead, they test how the fixed central scenario supply capacities and infrastructure respond to different levels of demand and to systematically higher or lower commodity and CO₂ prices (see section 10.1 and Table 26).