Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Cologne: groundwater extraction method probed

Investigations into the collapse earlier this month of the six storey archive building in Cologne this week focused on groundwater extraction.

It has emerged that contractors exceeded authorised dewatering limits while excavating a section of crossover tunnel for the city’s subway.

Cologne City Authority investigators said that groundwater pumping logbooks showed that main contractor Bilfinger Berger/Weiss & Freytag/Züblin joint venture, had built 15 extraction wells when it only had permission to build four.

The records also showed that from December last year this had nearly doubled the amount it was allowed to pump from 450m3/hour to 750m3/h.

A Bilfinger Berger spokesman said that the JV could not discuss the scheme’s construction methods.

The archive building collapsed on 3 March. Two people died and Cologne police said the collapse was caused by the ground under the building slipping into the excavation.

The joint venture was building the cut and cover crossover to allow light rail trains to switch tracks on the new Nord-Sud Stadtbahn Köln. The cavern site is in front of the building which collapsed and is just 300m from the river Rhine.

A Cologne Transit Authority (KVB) spokesman said work on the 28m deep excavation was at full depth or nearing full depth when the building collapsed. The 1m thick diaphragm walls supporting the excavation were supported by ground anchors in predominantly river gravel.

Groundwater level in Cologne is about 12m below the surface. Geotechnical experts said if the ground was more porous than initially thought, the volume of groundwater that had to be removed could have been greater than expected.

“[The amount of overpumping] means that there was far more water coming into the excavation than they expected and they were probably struggling to control it,” said HPR director Scott Steedman.

“There could have been underground channels or voids that were producing the flow.”

The transportation of fine material out of the soil matrix during pumping along with the high groundwater velocities could have combined to cause soil erosion and the formation of sink holes behind the diaphragm wall.

“If they were over-pumping the water they could be pumping fines – probably silt and fine sands – with the water and basically causing internal erosion in the material behind the wall,” said Card Geotechnics director Nick Langdon.

“You could wash out the fine gravel and silt in patches to leave clays and gravels [if present] that are not as easily transported without the matrix that held them together.

“That in itself would not directly cause the collapse but if this internal erosion happened near an anchor then the anchor would lose strength and ‘unzip’, closely followed by a few more and then of course the wall follows.

“The key will be the physical interrelationship of the dewatering points and the anchors if this is indeed the mechanism.”

A KVB spokesman said there were measuring devices on the diaphragm walls which did not show anything unusual immediately before the accident. Groundwater experts said the removal of fines should have been monitored and rectified.

“The flow of water carrying fines into wells can accumulate to the point of mining but that would be noticed early on,” said WJ Groundwater chairman professor Jim White.

“You would monitor for fines and if they [excessive amounts in groundwater extractions] were found to be a problem, the well would be turned off. [Overpumping] would be no reason to cause a catastrophic collapse, it would be more likely to be a dewatering failure.”

Experts also suggested that the increase in groundwater flow and corresponding pumping could have increased the hydraulic gradient under the excavation to the point where the ground liquefied, losing the passive resistance of the toe of the retaining wall.

“The hydraulic gradient [pressure loss/distance] would be at its steepest due to the confined space. If the critical hydraulic gradient or velocity is reached, the soil fluidises.

“The diaphragm walls would have a great deal of restraint from passive resistance of the toe. “If it was taken away, the wall would fail,” said independent consultant Andrew Hawkes/

Cologne mayor Fritz Schamma has ordered KVB and the environment department to give a full explanation of all aspects of the dewatering this week.

Readers' comments (5)

  • With sorrow at the loss of the archives and the two lives, I read this article with interest.

    The drawing in the article is really excellent and gives me some idea of how the incient occurreed - after all, it is not every day that a substantial 6-storey building largely diappears in seconds. Imagine a department store in your local town disappearing.

    I am surprised to discover that this modern building (1971 I believe) appears not to have been steel-framed - I cannot see anything substantial in the rubble which appears to be little more than concrete dust.

    The complexities of underground hydraulics are beyond my knoiwledge, however it does appear from tyhe article that the maximum extraction rate was specified - and exceeded by the contractors. Liquefaction is mentioned - is this a similar effect to that experienced by violent earth shaking during earthquakes ?

    An excellent article; informative to the layman as well as to the expert reader.

    Thank you - it has helped me greatly to understand the likely mechanism of this failure.

    Unsuitable or offensive? Report this comment

  • Being a civil engineer specialised in geotechnics I'd like to express my respect. Compared to the conspiracy theories and smattering German journalist provided during the last two weeks the present article is of an astounding expertise.

    Unsuitable or offensive? Report this comment

  • Jim Butterworth - not liqufaction as a phenomenon in some soils subject earthquake vibration, but the 'quicksand' effect when upward drag forces on soil particles from water flowing upwards added to the bouyancy forces together exceed the weight force of the particles.

    Unsuitable or offensive? Report this comment

  • Thank you Thomas - your helpful answer shows the complex conditions that need to be evaluated and quantified in such difficult construction areas. We live and learn (a little !).

    At the other end of the scale, I was closely involved in building a small swimming pool - when we came to excavate we found the soil was almost solid as there were no worms on site (don't know the reason for this). We had to get in a digger three times the weight just to shift the soil !

    Unsuitable or offensive? Report this comment

  • Le corbeau, ses yeux sur la coupe rivés,
    Nota une fiche "allégée"
    Le renard, par le gradient alléché,
    Ne demandait qu’à se déclencher...
    Mais les fines, jusque là suffisantes,
    Une petite cohésion pourvoyaient...
    Pauvres fines, par le courant entrainées,
    Ne purent plus leur fonction assurer !
    Le renard, enfin libéré,
    De la fiche diminua la butée.
    Les tirants, par ce défaut surchargés,
    A leur tour ne purent plus résister...
    A ces mots, le corbeau en conclut
    Que de Cologne l'Eau ne sent pas toujours bon,
    Et que le concepteur, pour ne pas l'avoir vu,
    Soit n'était pas meilleur, soit en avait trop bu !

    Unsuitable or offensive? Report this comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.