227,000 euros cheaper and still beaten: ARCEA challenges every score in the design contract for the Walloon collection centres at Malmaison and Obourg, finds one genuine error — but it could never close the 10.5-point gap to Greisch
The Council of State dismisses the action by the economic interest grouping ARCEA against the award of the design, inspection and safety-coordination contract for the Malmaison and Obourg collection centres to the engineering firm Greisch: the plea on undisclosed weighting and sub-criteria is inadmissible because ARCEA fails to link the rules it invokes to its complaints, the marks for a capped number of site meetings, a brief methodology and unmentioned design software are reasonable, and the only error the Council finds — the reproach that ARCEA would not supply ‘bordereaux’ — is worth at most 7.5 points against a gap of 10.5.
What happened?
The Walloon Region launched a general call for tenders for a services contract for the construction of two collection centres: at Malmaison, in the municipalities of Hensies and Bernissart, and at Obourg, in Mons. A single contractor was to fill three roles: design office for the design and supervision of the works (civil engineering, buildings and electromechanics), inspection office, and level-A health and safety coordinator. Specifications no. 221-05-E44 listed five award criteria in decreasing order of importance, without weighting: the detailed content of the services, the working method with staff and equipment, the time for completing the first phase, the price (six separate amounts) and the quality of the file. The contract was advertised in the Bulletin der Aanbestedingen of 18 November 2005 and estimated at 204,209 euros including VAT. At the opening on 30 November 2005 there were two tenders: from the economic interest grouping ARCEA for 329,222.85 euros and from the engineering firm Greisch for 556,600 euros, both including VAT. The services of the Walloon Ministry of Equipment and Transport (MET) found both tenders regular and the prices normal ‘given the specific nature of the services’. They broke the five criteria down into a table with sub-criteria, a distribution of points per role (design office, inspection office, coordinator) and a scale from ‘very poor’ to ‘very good’. ARCEA won comfortably on price, but lost on each of the four other criteria. For the design office’s ‘calculation notes, plans and documents’ it received 3.75 out of 15 against 11.25 for Greisch, with the remark that it did not specify that it would supply the ‘bordereaux’, formwork and reinforcement drawings and schedule, and that the number of meetings and visits was ‘apparently limited’; ARCEA had committed to an average of one meeting a week within a total geared to the estimated duration of the works, whereas Greisch set no limit. Its methodology was ‘not very detailed’ (3 out of 6), its staff according to the table included eight civil engineers ‘but not in construction’ (1.5 out of 6), it mentioned no design software for reinforced-concrete and steel structures (1 out of 3), and its selection file was ‘unclear and unstructured’ (0 out of 1). Final score: Greisch 81.3 points, ARCEA 70.8. On 31 March 2006 the Walloon Minister for the Budget, Finance, Equipment and Property awarded the contract to Greisch. ARCEA requested the award report on 30 June 2006, received it on 6 July and brought an action for annulment on 4 August 2006. The Region first argued that the grouping’s decision to go to court, taken on 24 July 2006, preceded the filing on 26 July 2006 with the registry of the Mons commercial court of the renewal of its directors’ mandates. The VIth chamber, presided over by Odile Daurmont, rejected that objection: the renewal had been validly decided in May 2006, the competent body had decided to litigate, and the fact that the renewal had not yet been published did not affect admissibility. The first plea, which complained that the weighting and sub-criteria had only been fixed after the opening of tenders and were never justified, did not even reach the merits: ARCEA listed a series of provisions and principles and, separately, a series of complaints, but drew no link between them. A plea must identify the rule breached and the way in which it was breached, ‘in principle in the manner of a syllogism’; this plea was inadmissible. In the second plea ARCEA challenged the scores one by one. The Council accepted that a bidder that caps its attendance at meetings and visits at an estimated total — and, as inspection office, its review at a maximum of three tenders — may receive fewer points than a competitor committing without limit. A brief methodology, where the specifications required a ‘very precise’ description per phase, justified 3 out of 6. The authority could score a team with higher qualifications more favourably without making a construction engineer a requirement, and a bidder that does not expressly state that it has design software can be marked down for it. On one point ARCEA was right: neither the report nor the specifications supported the reading that the Region later gave to the reproach about the ‘bordereaux’. But it had no interest in that finding. Even without the error it could not have scored more than Greisch on this sub-criterion, because its limited number of meetings remained a weakness and Greisch had other strengths. The impact therefore remained below 7.5 points, while the gap was 10.5 points. The action was dismissed; ARCEA bears the costs of 175 euros.
Why does this matter?
The judgment is a lesson in how not to build a challenge to a qualitative assessment. ARCEA had an interesting point: the specifications gave only an order of importance, and the weighting, sub-criteria and distribution of points per role first appeared in the table attached to the award report. Whether that was permissible the Council never examined, because the plea set a list of rules alongside a list of complaints without connecting them. The second part shows the limits of a score-by-score attack. The Council reviews marginally: a score is unlawful only if its reasoning is incorrect or manifestly unreasonable, and even an established error leads to annulment only if it could overturn the ranking. The Council did the arithmetic itself here: on the 15 points of the sub-criterion, the error could shift at most 7.5 points, against a gap of 10.5 points. On substance, the judgment is instructive about what a tender ‘commits’ to. A bidder that caps the number of site meetings or tenders to be reviewed in a services bid builds in a weakness against a competitor committing without limit. And what seems obvious — a design office surely has design software, execution drawings surely include reinforcement drawings — does not count if it is not in the tender.
The lesson
For bidders: state expressly in a qualitative tender what you will deliver and what you have at your disposal, including what you consider self-evident; the evaluator scores the text, not your reputation. Be careful with caps on meetings, visits or reviews: they protect your margin, but cost points if a competitor commits without limit. If you want to challenge an assessment, link each complaint to a specific rule of law, and work out beforehand whether the errors you identify could together bridge the points gap; if not, you have no interest in the plea. For contracting authorities: a table with short, concrete findings per sub-criterion can suffice as reasoning, but each criticism must be traceable in the tender; explaining it differently afterwards in a brief does not save an incorrect reason. Preferably set out the weighting and sub-criteria in the specifications, so that the debate on late weighting cannot arise at all.
Ask yourself
As a bidder, does your tender expressly mention all the plans, documents, software and commitments the specifications require for each criterion, including what is obvious to you? Have you assessed how many points a cap on meetings, visits or reviews may cost you against an unlimited commitment? Does your application link each complaint to a specific rule of law and to the way in which it was breached? Have you checked whether the assessment errors you allege could really close the points gap with the selected bidder? As a contracting authority, is every negative reason in your evaluation table based on something that is literally in, or missing from, the tender? As a contracting authority, have you set out the weighting and sub-criteria in the specifications, rather than developing them only at the evaluation stage?
About this database
The Council of State (Raad van State / Conseil d'État) is Belgium's supreme administrative court. In disputes over public procurement — from contract awards to tenderer exclusions — the Council of State is the final arbiter. The rulings in this database are summarised by TenderWolf in plain language, with practical lessons for tenderers and contracting authorities. View all rulings →