Horizon Europe Lump-Sum Budgets: The Detailed Budget Table, Worked Example

About half of the 2026-27 Horizon Europe call budgets are now paid as lump sums. For a coordinator that changes when the hard budgeting work happens: instead of reporting actual costs for four years, the consortium estimates them once, in detail, before submission, and the resulting amounts are fixed per beneficiary and per work package in the grant agreement. This page explains the mechanics and then builds a complete budget for SALTMARSH, the fictional eight-partner, 42-month project whose work plan is laid out in the work packages and Gantt worked example.

What a lump-sum grant is, and is not

A lump-sum grant sets one share of the lump sum for each work package. When the work package is completed, that share is paid. The Commission’s guide is explicit about what completion means: “The completion of work packages is not based on a successful outcome, but on the completion of activities as described in the description of action.” A failed experiment that was carried out as planned is a completed work package.

Everything else about the proposal stays the same. You use the standard application form, you address the same three evaluation criteria, and the page limit rises only slightly, to 45 pages for RIA and IA topics, with section 3.1 given an indicative 17 pages instead of 12 so there is room for the fuller work plan. What disappears is actual-cost reporting: there are no financial checks or audits by the Commission, and no timesheets, payslips or invoices to show the EU. Technical reviews, and checks on IP, ethics, open science and dissemination obligations, still apply.

Type 1 and Type 2

The topic text states which of the two lump-sum options applies. They differ in who sets the amount and therefore in what you have to submit.

The detailed budget table, sheet by sheet

For Type 2 topics you download the Excel file from the submission system, fill it, and upload it with Part B. Cost estimations must be in line with the beneficiary’s normal practices, reasonable, in line with the proposed activities, and subject to the normal eligibility rules: a cost can be estimated only if the same type of cost would be eligible in an actual-cost grant.

One change is under way. During 2026 the Commission is progressively replacing the Excel file with an online detailed budget table built into the Part A forms, and the two systems run in parallel. For the first topics on the online table, HORIZON-CL5-2026-05 topics D5-07, D5-02, D5-12 and D5-13 and HORIZON-CL5-2026-07 topics D1-01 to D1-05, there is no Excel template in the submission system at all. The level of detail and the evaluation approach are unchanged, so everything below applies to both.

Personnel costs and the Horizon dashboard

Personnel is usually the largest line, and it is the one evaluators check most directly. They are told to use the Horizon dashboard for lump-sum evaluations as an orientation: for each combination of country and organisation type it shows the distribution of average monthly personnel costs between the 20th and the 80th percentile, based on grants already signed in Horizon Europe. A rate above that band is not forbidden, but it must be explained in the Any comments tab, and if the budget as a whole is unfit for purpose or strongly overestimated the Implementation score drops. Consultants commonly advise treating anything above the 80th percentile as needing an explicit justification.

The person-month itself is simple: one person working full time for one month, and in the table one person-month is one item. The glossary entry on person-months covers the working-day conventions, and the budget calculator converts person-months to costs per partner.

[Beneficiary short name], personnel category [A1 employees]:
The average monthly cost of [€X] per person-month is above the Horizon dashboard
range for [organisation type] in [country]. It reflects [the salary scale set by
the national collective agreement / the seniority mix: N senior researchers
(grade) and N postdocs] required for [task numbers], and includes [employer
social contributions of Y% required by law]. The figure is the average of the
actual monthly costs of the named staff in [year], consistent with the
beneficiary's usual cost accounting practice.

Worked example: the SALTMARSH budget

SALTMARSH is a Research and Innovation Action at a 100 percent funding rate: eight beneficiaries, six work packages, 424 person-months over 42 months. Its work plan, deliverables and staff-effort matrix are in the companion worked example; here the same person-months become money.

Step 1: personnel rates

Step 2: the other cost categories

Two beneficiaries subcontract: DELTA buys €120,000 of marine works for the pilot installations in WP4, and BWT buys €30,000 of video production in WP6. Both tasks are support work, not core research, and both are described in table 3.1g. Purchase costs cover travel to meetings and pilot sites, sensors and pilot materials, events and open-access fees. DELTA also needs a hydrodynamic test rig for the design work in WP3, and here the depreciation rule matters: only the share used by the project is eligible.

Step 3: the checks evaluators will run

Two checks fall straight out of the table. The first is the 15 percent rule: any beneficiary whose purchase costs exceed 15 percent of its personnel costs must complete table 3.1h in Part B. In SALTMARSH that is five of eight beneficiaries, mostly the pilot hosts buying sensors and materials, which is normal for a project with physical pilots and exactly what 3.1h exists to explain.

The second is balance. No country holds more than 23.5 percent of the budget (the Netherlands, driven by DELTA’s engineering role), comfortably below the 40 percent level that consultants treat as a sign of an unbalanced consortium; that threshold is a heuristic, not a Commission rule. Subcontracting is 4 percent of the total, well under the 15 to 20 percent that the same consultants advise as a ceiling.

Step 4: the lump-sum breakdown

This is the table that matters most, because it becomes Annex 2 of the grant agreement: one amount per beneficiary per work package. The Excel sheet generates it; the Part A budget table is then filled from its beneficiary totals.

Splitting long work packages by reporting period

A work package is paid at the end of the reporting period in which it is declared complete. In SALTMARSH the reporting periods are months 1 to 18, 19 to 36 and 37 to 42. Management (WP1) and dissemination (WP6) run the full 42 months, so as single work packages they would only be paid at the very end. The Commission’s guide says work packages with a long duration may be split along the reporting periods, with the same content, precisely so that activities can be declared and paid earlier; each split work package then needs its own description in table 3.1b. Splitting WP1 and WP6 pro rata to months turns six work packages into ten.

The trade-off is more tables in Part B and more work packages to declare, in exchange for interim payments that follow the work. Without the split, the RP1 payment would fall from €756,000 to the €499,500 of WP2 alone. The heavy RP2 payment is a consequence of the pilot design, not of the split: WP3 and WP4 together are more than half the budget and both end in the second period. The Mutual Insurance Mechanism and reporting period are defined in the glossary.

What happens when a work package is not completed

At the end of each reporting period the coordinator declares each work package Completed or Not Completed, supported by the technical periodic report, and the financial statement is generated automatically from the accepted work packages. The Commission’s own example shows the arithmetic: four beneficiaries, five work packages worth €350,000, €470,000, €350,000, €200,000 and €300,000, of which the first and third are completed in the period.

• An incomplete work package is simply paid later, at the end of any subsequent reporting period in which it is completed.
• A work package can be accepted when all essential tasks were completed, when equivalent tasks were carried out, or when deviations are justified. Amendments to reflect scientific or technical changes are also possible.
• Before a work package you declared completed is rejected, you are invited to respond to the project officer’s observations. If the rejection stands, that share is not paid at that point.
• Only at the final reporting period can a work package be declared Partially Completed with a percentage. If it cannot be finished, for technical reasons or force majeure, the lump sum is paid partially in line with the degree of completion, decided case by case.

The practical lesson for design is to keep every work package completable by the consortium itself. A work package that can only be declared complete once an external authority grants a permit is a payment risk; put the permitting activity in its own task with a mitigation in table 3.1e, and define completion as the activities carried out.

From evaluation to grant agreement

The no-negotiation principle applies: the grant agreement is prepared from the proposal as submitted, with only obvious errors corrected, changes required by the rules, and the lump sum adjusted to the amount in the Evaluation Result Letter. The breakdown per beneficiary and per work package becomes Annex 2; the detailed budget table you submitted is not part of the grant agreement. Once fixed, the lump sum is not questioned if prices for goods or services change. Inside the consortium you may use the money as you see fit, since the Commission does not see the actual distribution, but moving shares between work packages in the grant agreement requires that the work packages are not yet completed, that the move is justified by the implementation, and an amendment. How the consortium agreement should handle that internally is covered in the consortium survival guide.

Eight mistakes evaluators flag

1. Personnel costs above the dashboard with no comment. Use the Any comments tab and the template above.
2. Core tasks subcontracted. Subcontracting is for support tasks, one line per beneficiary per work package, each justified in table 3.1g.
3. Work packages whose completion depends on third parties. Permits, regulators or recruitment outside your control make the share unpayable.
4. One management work package spanning the whole project. It will not be paid until the end; split it along the reporting periods.
5. Equipment at full price. Enter depreciation for the share and period used, unless the work programme explicitly allows full capitalised costs.
6. Purchase costs above 15% of personnel with no table 3.1h. The rule applies per beneficiary.
7. A budget concentrated in one country. Consultants flag more than 40% as a sign of imbalance; it is a heuristic, but evaluators notice.
8. Part A not matching the Excel file. Copy the beneficiary totals from the estimated breakdown of the lump sum, and keep table 3.1f consistent with the person-months overview.

For the Part B side of these checks, see the Part B template walkthrough; for writing the narrative that justifies the numbers, see budget narrative mastery and budget justification. On whether lump-sum topics are less competitive, an earlier Commission assessment reported higher success rates, discussed in Horizon Europe’s hidden instruments; treat it as indicative rather than a current programme-wide figure. For quick per-partner estimates before you open the Excel file, try the quick budget calculator.