33 x 48 x 81 cm with 10% tolerance: a coffee machine ten centimetres too big does affect the price and maintenance criteria
Liège's Citadelle Hospital awarded a five-year framework for coffee and 90 coffee machines to Illico, but the Council of State suspends the award because it is unclear which machine model Illico actually offered, and the tested model exceeded the bespoke dimensions by ten to twenty per cent with no reasoning given for that deviation.
What happened?
The Centre Hospitalier Régional de la Citadelle in Liège launched on 2 April 2021 a framework for the supply of freeze-dried coffee with provision of new coffee machines (installation and full-omnium maintenance included) over 60 months. Minimum: 7 million doses of 12cl over the term; maximum: 10 million doses. The tender received seven bids, including market leader Jacobs Douwe Egberts Pro Be and eventual winner Illico. Technical clauses imposed machine dimensions: 33 cm wide, 48 cm deep, 81 cm high, 'with a tolerance of 10%'. Bidders had to deposit two test machines (one for nursing units/kitchen with a minimum 1-litre thermos option, one for 'other destinations') and could offer only a single model for both destinations. On 6 May 2022 the CHR executive awarded to Illico, based on an analysis report of 21 February 2022. Ranking: Illico first, JDE second. JDE filed an extreme-urgency action on 1 June 2022, represented by attorneys Virginie Dor, Flore Verhoeven and Mathilde Vilain XIIII, with two pleas: (1) Illico's machine exceeded the 10% dimension margin — substantial irregularity; (2) Illico did not meet the turnover selection criterion. The hospital first contested JDE's standing — the Council brushes this aside: as second-ranked JDE plainly has standing. Then the substance. The Council identifies a confusing problem: which model did Illico actually offer? The test machine deposited turned out to be the 'OptiVend s11 NG' — but it is not certain that this was the model in the bid. On the eve of the hearing the CHR admitted another sub-model ('OptiVend TS NG', thermos version) might have been offered. The Council reads the tender strictly: 'A clear text is not to be interpreted, even if it does not reflect its author's intent, lest third parties be misled'. A 10% tolerance means deviations beyond that are inadmissible — in either direction. The OptiVend s11 NG exceeds width and height by over 10%. Even with 'door and lid open', depth still exceeds by more than 10%. The TS NG alternative also fails: in both configurations the dimensions deviate by more than 10%. The CHR's defence that 'dimensions have no effect on the award criteria' is dismissed: (a) smaller machines are considerably cheaper with different use and maintenance characteristics, and (b) the tender expressly states that the unit price 'includes provision with installation and full-omnium maintenance', and Illico itself confirms in its price justification that the choice of machine is 'significant' for the price level. The fourth award criterion (maintenance) also shows noticeable differences between models. Neither the analysis report nor the award decision explains why non-compliant dimensions were not a problem — breach of the verification and reasoning duties under Article 83 of the 2016 act and Article 76 of the 2017 Royal Decree. Whenever the authority is confronted with a difficulty, it must 'articulate the reasons why it nonetheless considers the bid regular'. The Council suspends at extreme urgency, with the balance of interests favouring suspension.
Why does this matter?
This ruling captures three hot points that converge in big 'consumables + infrastructure' contracts (think: coffee + machines, office supplies + printers, food + kitchen equipment). One: when the tender imposes dimensions with an explicit tolerance, those limits are binding in both directions — and they rarely only concern spatial fit; they almost always influence price and maintenance characteristics, directly touching award criteria. Two: the model offered must be unambiguously identifiable from both the bid and the test specimen. Testing an 'OptiVend s11 NG' while the bid refers to an 'OptiVend TS NG' (or vice versa) is a structural verification problem. Three: a positive regularity decision may be motivated more briefly than a rejection, but the moment the authority is confronted with a 'difficulty' (here: a visible dimension deviation), it must state in its decision why it has accepted the bid despite that difficulty. This is a sharpening relative to the standard: you need not comment each regular bid individually, but you cannot sweep an obvious problem under the rug.
The lesson
If you lose to a bid where you suspect the product physically does not meet a tender specification (dimensions, capacity, power) with an explicit tolerance: compare the technical datasheet in the bid with the deposited test specimen and with publicly available catalogue data. A discrepancy between these three sources is on its own enough to construct a serious plea — the authority's verification duty then becomes problematic.
Ask yourself
The winning bid involves material you yourself know, and you suspect the model falls outside the tender tolerance: (1) check the technical datasheet in the bid; (2) check the deposited test specimen; (3) check the manufacturer's catalogue. If these three do not align, you have a serious plea.
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 →