What we committed, what we retained, what we deferred

An 80MW solar proposal used expected annual output of 160GWh and a EUR72/MWh annual market-price assumption. The multiplication produced EUR11.52 million revenue before other costs. The development team wanted to use that figure in the investment paper because it was consistent with a familiar market reference. I asked whether it represented the prices likely to coincide with the asset's actual production profile.

Technical and commercial colleagues examined the hourly assumptions and contract possibilities. The generation-weighted price in the revised uncontracted scenario was EUR54/MWh, implying EUR8.64 million on the same expected output. The EUR2.88 million difference was a change in the planning account, not an observed loss or a reduction in physical generation. It showed that the original average and the asset's forecast output did not belong together in the way the paper had assumed.

We then assessed a contract covering 25 per cent of actual generation at EUR62/MWh, with the remaining output exposed to the modelled market profile. On expected annual generation, that meant 40GWh at the fixed price and 120GWh at the EUR54/MWh assumption, giving EUR8.96 million before other costs. The arrangement was based on the specified share of actual output, not a promise to deliver a fixed forecast volume regardless of weather. Commercial and legal colleagues reviewed its detailed terms.

The board deferred the construction commitment while the project was reworked around that revenue view and its financing implications. The contract arrangement was agreed subject to the project's stated development conditions. We did not report EUR8.96 million as earned revenue or imply that the fixed-price portion removed all project risk. Output, curtailment, costs and the remaining market exposure still mattered.

The development team accepted additional work and a slower approval timetable rather than retain the higher opening figure. My contribution was to ask whether the price assumption described the energy we expected to produce. A more detailed forecast is still uncertain, but it should at least be built from components that can sensibly occur together.

I would put the generation profile beside the proposed revenue assumptions at the start of appraisal, before an annual energy total gives the investment a misleading simplicity.

At a March 2025 review, consolidated cash was EUR42 million and shareholders proposed a EUR24 million distribution. The total made that request appear possible with a substantial balance left over. Treasury's entity-level account showed a different position: EUR30 million was held in project companies and EUR12 million at the parent. The project money carried obligations that could not be understood from the consolidated figure alone.

Within the EUR30 million, EUR12 million supported debt reserves and EUR9 million was committed to capital work. The remaining EUR9 million required confirmation of the conditions for movement from the relevant project companies. Finance, treasury and legal colleagues reviewed those conditions before treating it as accessible to the group. We did not assume that a positive project balance made every euro distributable or that a contractual reserve was surplus simply because no immediate payment was scheduled.

Some shareholder representatives preferred the original proposal, arguing that the group should not retain unnecessary liquidity. I agreed with the principle but asked the board to distinguish cash retained for defined obligations from cash held without a purpose. After the review, EUR21 million was available across the parent and the eligible project balances. A EUR24 million payment would have exceeded that amount before allowing for a continuing liquidity cushion.

The board approved and paid EUR12 million after the required entity-level confirmations. EUR9 million of accessible liquidity remained, alongside the EUR21 million supporting reserves and committed work. The lower payment did not change the group's total earnings, and we did not present it as value created by finance. It changed the timing of shareholder cash while preserving resources the operating plan still required.

The case improved how we prepare distribution papers. A consolidated balance now arrives with an account of location, purpose and movement conditions. The board can question whether those conditions or reserves remain appropriate through the proper process. It should not, however, make a payment recommendation by treating all reported cash as interchangeable before that work has been done.

A EUR120 million refinancing proposal offered a lower opening rate and a longer maturity. The rate comparison moved from 5.9 per cent to 5.1 per cent, while the remaining term moved from three years to seven. The initial commercial summary emphasised the 0.8 percentage-point reduction. Applying it to an unchanged EUR120 million balance suggested EUR960,000 of annual interest reduction. I asked for the payment schedule and transaction costs to sit beside that attractive opening calculation.

Fees totalled EUR1.8 million. Treasury modelled the repayment profile and the assumptions needed to compare the old and new arrangements consistently. Legal advisers reviewed the conditions, and the board considered the longer funding horizon separately from the interest benefit. The comparison did not assume that the company could extend maturity without accepting continuing obligations or that the opening balance would remain constant throughout the term.

Some directors questioned paying a substantial amount upfront when the reported annual interest benefit was smaller. Others valued the additional time before refinancing would again be required. We presented those as distinct considerations. A simple fee-to-opening-interest ratio would not establish the full economic payback because balances, repayments and other conditions changed over time. The board approved the arrangement with both its funding purpose and near-term cash cost explicit.

The transaction completed in 2024. Over the first twelve months, the interest reduction on the comparable balance schedule was EUR870,000 rather than the EUR960,000 constant-balance illustration. Against EUR1.8 million fees, that meant the first-year interest benefit had not recovered the transaction cost. The net difference on those two items was EUR930,000 of additional cost, before considering the wider funding consequences. We did not describe the refinancing as immediate net cash saving.

My contribution was to make the recommendation robust enough to survive the less flattering first-year account. A funding decision can be worthwhile for its maturity and longer-term economics while having an initial cost. Directors should approve that exchange knowingly, rather than discover later that a lower percentage was only one part of the obligation they accepted.

I would compare refinancing options on the first-year cash requirement before discussing the headline rate, making fees visible while the negotiating alternatives are still open.

Storm damage at an operating asset in 2023 led to a EUR3.6 million insurance claim. The operating team needed to progress the authorised recovery work while the claim was assessed. An early cash forecast assumed the full amount would arrive promptly. I asked treasury to prepare a funding view that did not depend on either that timing or full admission of the claim. Insurance and legal specialists continued to pursue the recovery on its own merits.

We arranged a EUR4 million temporary facility under the company's approved funding process. Peak borrowing was EUR1.5 million while receipts remained uncertain. The operating programme and its technical requirements were determined by the responsible asset leaders; finance did not use the claim forecast to decide what repair work was necessary. The facility provided a way to meet the approved commitments without treating a submitted amount as money already available.

The arrangement had a cost. Interest was EUR55,000 over the period and arrangement and availability fees were EUR50,000, giving EUR105,000 in total. Some colleagues preferred to rely on the expected insurance receipt and avoid that expense. The board accepted the funding cost after considering a slower or smaller recovery. We did not present the facility as an operating improvement or as evidence that the insurer would ultimately pay the submitted amount.

EUR2.4 million was admitted and received nine months after submission, in 2024. The temporary borrowing was repaid from the available receipts. EUR1.2 million remained disputed at the review point and was not included as collected cash. Reporting colleagues assessed the relevant accounting separately; the liquidity decision did not establish final entitlement or the appropriate carrying value of an unresolved claim.

The lesson was about the relationship between a specialist process and a cash plan. The company could pursue a supported claim firmly while funding its obligations without assuming the desired answer. I wanted directors to see the price of that independence and the uncertainty it addressed, rather than allow a favourable receipt date to make the forecast appear comfortable until it failed to occur.

For a future storm scenario, I would distinguish recovery spending from possible insurance receipts in the first liquidity exercise, rather than let one assumed date connect obligations with different owners.

A proposed 40MW, 80MWh battery investment came with EUR28 million capital expenditure and an attractive annual revenue presentation. The paper combined EUR4.1 million from energy trading with EUR2.7 million from system services, producing EUR6.8 million. Each estimate had been prepared using assumptions suited to its own activity. I asked whether the same equipment could deliver both sets of results within one feasible operating schedule.

Engineering and commercial specialists reviewed the combined dispatch, power and energy limits, service requirements and degradation assumptions. The concern was not that multiple revenue streams were impossible. It was that separately attractive schedules might require the same capacity at incompatible times. Their joint assessment produced EUR4.9 million of modelled annual revenue under the revised assumptions, rather than the sum of the two standalone estimates.

The revised view included EUR1.2 million annual operating and market-access costs, leaving EUR3.7 million before financing, tax and the further lifecycle allowances still being developed. That amount was not labelled free cash flow or a completed investment return. The connection position, future replacement needs and sensitivity to market conditions required additional work. Finance's role was to make those incomplete parts visible without substituting for technical dispatch analysis.

The development team wanted to retain the timetable and argued that the project could still have a valuable role beside existing generation. The board agreed that the opportunity deserved further examination but withheld the full construction approval. It authorised EUR240,000 for dispatch and connection studies, with a defined return to the investment discussion. The studies were in progress at the point described; no operating revenue or completed asset is claimed.

We gave up an immediate commitment and accepted further development cost to obtain a more credible account. The EUR1.9 million reduction between models was not an actual revenue loss, and the unapproved EUR28 million was not a saving. The decision established what evidence the next capital request would need. One asset may earn several forms of value, but its board should see how the physical and commercial obligations fit together before adding the numbers.

I would have the operating team test compatibility between proposed battery revenue streams before finance consolidates them, preventing a detailed model from lending authority to mutually competing uses.