Design Risk Assessment isn't current best practice. It should never have been - the CDM regs don't require it. DRAs of a severity x likelihood type have never been suited to the design process, though lots of CDM-Cs are unable to grasp that.
The HSE at least does get it - see
http://www.hse.gov.uk/co...on/cdm/faq/designers.htm - "The consideration of hazard and risk is integrated within the design process, so there is no need to carry out a separate 'design risk assessment'."
Or see the CITB guidance at
http://www.cskills.org/u...rs4web_07_tcm17-4643.pdf which was produced and published with the HSE and supported by a wide range of industry bodies including the ICE and IStructE. This states "The introduction of CDM 2007 allows the process of 'risk assessment' to be revisited with a view to adopting a more productive and meaningful approach" and proceeds to describe a 'hazard elimination and risk reduction' process. The CITB guidance states that it "illustrates the necessary
sequence of actions … compatible with the need for a simple qualitative technique. It will only be on rare occasions (for example in the design of nuclear facilities) that a more sophisticated approach, such as quantitative analysis or risk ranking, will be required".
The CITB guidance includes a template for recording the process and outputs. The information it recommends be recorded is:
1. References/Location/Phase etc.
2. Hazard or issue - Brief description (if helpful, these can be wider project hazards,
e.g. planning issues).
3. Elimination? - If 'yes' describe action to be taken, by whom and by when.
4. (If elimination not possible) Associated risks and reduction measures taken - Note:
action to be taken in accordance with the hierarchy and principles of prevention and
protection
5. Information to be conveyed to others - Is the method chosen to communicate this
information adequate?
6. Date of assessment
7. Action by and Required by:
8. Action cleared
There is no attempt to rank or rate hazards or risks is included in this list. Risk assessments that score (numerically or verbally) the probability and/or severity of an occurrence or some combination of the two are not current best practice, and should not be used unless the scoring is properly undertaken with a detailed, calibrated, quantitative analysis.