BIM: Myth of Clash Detection

The myth

‘Clash Detection’ is one of my favourite subjects, in a cranky, gets me annoyed kind of way. It has become the automatic response to justify the benefits of BIM by those who don’t understand the design process.

If not used properly, ‘clash detection’ will make the the design coordination process more difficult, more time consuming and will produce an inferior design result.

Clash detection is fantastic for final validation that the design is correct- but is only useful only once you have gone through the right design process.

Technology over commonsense

I see more than a few projects where the various design consultants or subcontractors will independently put a great deal of effort into modelling their design without much planning, communication or forethought.

The models are then federated by the BIM Manager in a ‘clash detection’ process to see if it fits together. (of course it won’t …)

The approach is generally ‘first-in, first served’, so arguments ensue over who needs to move their beam/pipe/duct/cable tray. The number of clashes can be gigantic (I’ve seen 40,000 in a single federated model) and individually resolving each clash is a slow, tedious or even painful process.

It can be a triumph of technology over commonsense.

It is much better to use BIM to plan and communicate, then model correctly the first time and use these clash tools to check it all works.

Good design: Much more than clash free

Clash detection is typically used for building services & structural design coordination. However, there is a lot more to a good design than the various components merely not clashing with each other. A good design should be:

>Logical and efficient.
For example, pipes and ducts in straight runs are obviously quicker and cheaper to construct than those snaking up, down & around.
>Repetitive where possible.
For example, a common approach to the arrangement of structure & building services or in sequence of construction trades.
>Readily constructable
It is most efficient if each building services trade is completed in a few major phases and each trade completed sequentially.
This is as opposed to trades being intertwined with each other, which requires either piecemeal work (multiple site visits or intermittent work) or several trade working concurrently (which slows down everyone)
>Readily maintained and serviced. Particularly in building services, many components have a finite life, or require regular maintenance. So the ability to readily replace or access certain components should be considered as part of the design, and BIM plays a part here in modelling access zones and planning future work. The UK PAS1192 document refer to this kind of modelling, but I rarely see it.

Therefore, a ‘clash free’ design is obviously necessary, but it is not the only criteria for a quality or complete design process.

Change in Behaviour

The strange thing is that the behaviour of design consultants, contractors & subcontractors (who really should know better & end up wasting their own time) has been influenced this capability of software, so that the design process is flipped on its head.

In other words, we are doing something not because it makes sense, but because we can.

A Design Issue not a BIM issue

An interesting side effect of this phenomenon is that work that previously (and rightly) would have been a design coordination issue has now become a BIM issue. In other words, the task of identifying and resolving these issues seems to become the responsibility of the project ‘BIM Manager’ or similar, rather than the designer.

Start big, work down to the detail


BIM can play an important part in the design coordination process, which culminates in ‘clash detection’.

However, a common mistake is that too much model detail is added too early, with not enough attention in the planning stage. BIM can play an important part in this planning stage:

  • Agree typical design zones
    for structural, building services or fitout/finishes. This obviously has an effect on the construction sequence. If necessary this could be modelled, as shown in the image above.
  • Develop a ‘design zone’ model
    including horizontal and vertical zones allocated to each building service, structure & so on. This can be a fairly simple massing model that just shows the overall anticipated extent. Refer to my video below.
  • Commence discipline modelling
    by staying within respective zones. The zones for maintenance/replacement access should be modelled as a massing ‘exclusion’ zone.
  • ‘Anti-clash’ checks
    By visually comparing the zone model and each discipline/trade model, the design can be quickly & easily checked. I guess you could see this as an ‘anti-clash’ i.e everything from a particular discipline/trade should fall completely within the zone. It is only where the model goes outside of the zone that you need to worry.
  • Deal with exceptions
    There will obviously need to be exceptions where services have to pass through another zones and these can be dealt with individually. This could also apply to suspension or support systems.
  • Clash detection
    Once this has been completed, then a clash detection process can be completed as a final validation to make sure nothing has been missed.