Dispersion and Ventilation
This section covers:
- dispersion of flammable or toxic gases or vapours;
- dispersion within a module, over an installation, over the sea-surface, and sub-sea;
- dispersion of releases which are passive, momentum-dominated or buoyancy-driven;
- ventilation of open, sheltered or enclosed areas, primarily as a means of minimising gas build-up and aiding dispersion;
- forced or natural ventilation; and
- ingress of smoke into the TR.
Dispersion and ventilation are recognised as key topics in the control of fire and explosion hazards on offshore installations. They have been the subject of various research studies over the past 10 years and are the subject of increasingly detailed modelling by dutyholders as a means of demonstrating lower risk levels.
- identify areas of uncertainty in evaluation of ventilation and dispersion.
- promote the use of consistent methodologies in the evaluation of installations ventilation and dispersion analysis,
- Initiate research to increase knowledge and understanding of ventilation and dispersion analyses,
Knowledge of dispersion and ventilation
Standards, guidance and codes of practice on ventilation
The standards, guidance and codes of practice relevant to ventilation on offshore installations include BS 5925, ISO15138, BS EN60079-10, IP 15, API 500 and NORSOK H-001. A review of these documents has highlighted a number of areas of concern:
- there is a lack of consistency between the various standards which apply to ventilation and dispersion in offshore modules. In particular, the criteria against which the adequacy of ventilation can be assessed are inconsistent, with some standards using air change rates as the key criteria, whilst others use ventilation rates;;
- the validity of the performance criteria in current standards is uncertain. In particular, measures related to air change rate are of doubtful validity, as this measure gives no indication of air distribution and is extremely difficult, if not impossible, to quantify in practice;;
- the specification of 12 air changes per hour as representing adequate ventilation in naturally-ventilated modules (IP15, API RP500) lacks a sound basis; and;
- the applicability of the current standards is unclear:
- the maximum leak size which is effectively dispersed by the recommended ventilation rates is not clear;
- no account is taken of the influence of the leak on local ventilation conditions;
- basing the definition of 'open' areas on wind speeds may not be the best measure; and
- if wind speeds are to be used to define open areas, the existing guidance should make it clear what location the speeds are referred to.
Research on gas dispersion
Notable research studies on dispersion on offshore installations include:
- Phase 1 of the BFETS JIP (Report OTI 92 591 [PDF 1mb]) the main conclusion of which was that, whilst much of the phenomena of gas releases is understood, there is much uncertainty in the way that these phenomena would interact in real offshore scenarios;
- JIP on 'Gas Build-Up from High Pressure Natural Gas Releases in Naturally Ventilated Offshore Modules' (Cleaver et al, 1998 and 1999), involving gas dispersion trials in the large scale test rig at Spadeadam. This project also included a CFD model evaluation exercise (showing good qualitative agreement) and development of a workbook approach for flammable gas cloud prediction;
- A study by Saunders et al (2001) involving on-site measurements and CFD modelling to determine the adequacy of natural ventilation on offshore installations; and;
- On-going work by HSL to review test methods for smoke/fire HVAC dampers.
[back to top]
Modelling and measurement capabilities
There are a number of areas of uncertainty with regard to the modelling of offshore ventilation and dispersion:
- It would appear that models for predicting the size and composition of gas clouds which could form in congested confined modules, have not been extensively validated. This is in part to a lack of reliable experimental data, which is now available (Cleaver et al, 1998, 1999);
- Models developed prior to the recent JIP data (Cleaver et al, 1998, 1999) should be validated against that data;;
- The main CFD model validation activity has only recently been published (Savvides et al, 2001a, b). The performance of two CFD codes; FLUENT and FLACS is claimed to be good when compared with large scale data, however the details of the validation exercise are not yet available;;
- An integrated CFD-based approach to the modelling of ventilation, gas dispersion and explosion simulation has become more commonplace. It is highly likely that compromises are being made in this approach, but the uncertainties have not yet been addressed;;
- The competency of those using CFD models is a key issue, but there is no 'best practice guidance' available for offshore applications;;
- The validity of models for ventilation and dispersion in congested confined modules in cases where the wind speed is low, is uncertain.
Measurement of ventilation rates in enclosed, mechanically-ventilated areas is relatively straightforward and well documented. Measurement of ventilation rates in relatively open conditions is much more difficult and depends on the flow regime. Under plug flow conditions air speed measurements or tracer gas techniques can be employed. The data would then have to be correlated with wind speed and direction. This approach can, however, be invalidated where 'short circuiting' of the air flow occurs. Smoke releases or point measurements of air velocity on a deck can highlight areas of recirculation or low wind speed, but do not provide any information on ventilation rate.
Wind tunnel modelling and a range of mathematical modelling techniques are often used to evaluate offshore air movement and dispersion. In a sample of recent Safety Cases only one provided validation of modelling against offshore experimental measurements, as required by ISO 15138.
Strategy development issues
Standards and guidance
- The current standards, guidance and codes of practice are inconsistent, in some cases lack validity, and their applicability is uncertain.
- There is a need to re-examine the basis of current standards and guidance, with a view to defining more appropriate and practical measures of performance.;
- Establish, in consultation with industry. The feasibility of drafting best practice guidelines for the application of CFD and other approaches (eg workbook approach) to the prediction of offshore ventilation, dispersion, and gas build-up, particularly with respect to the effectiveness of gas and leak detection systems.
- A recent JIP has provided much-needed data on gas cloud build-up in offshore modules (Cleaver et al, 1998, 1999; Savvides et al 2001 (a), (b)). Although a total of 66 tests were undertaken only two of these were at wind speeds of less than 1.5 m/s. Empirical models which have been devised using this data (Cleaver & Britter, 2001) could give misleading results if applied outside their range of applicability, for example in conditions of low wind speed. The performance of more advanced modelling approaches, such as CFD, is also uncertain for conditions outside those investigated in this JIP;
- It is not clear whether models developed prior to the recent JIP data (Cleaver et al, 1998, 1999), such as CHAOS, have now been validated against that data.;
- CFD is being increasingly used to predict gas dispersion and build-up, prior to explosion simulations. It appears that a range of modelling approaches are being employed, for example coarse grid simulations in one case (Holen, 2001), through to the use of half to one million grid cells in another (Savvides et al, 2001 (b)). Best practice has, it seems, not been established or agreed. Neither has the sensitivity of predictions of gas cloud size to the CFD modelling approaches typically being employed by industry been established.
Dispersion calculations in exceedance modelling approaches
- A range of predictive techniques are being employed to calculate gas build-up and dispersion in exceedance modelling approaches. Since explosion over-pressure is related to gas cloud composition and size, it is important to understand that uncertainties in methods used to predict offshore ventilation and dispersion will ultimately be reflected in calculated exceedance curves.;
- Guidance is need on the number of data points required to obtain statistical validity of an exceedance curve, and the methods used to interpolate and extrapolate from limited data points. In particular the validity of the 'frozen cloud' and 'symmetry' approximations used to extrapolate from a few hundred CFD simulations need examination.;
- Develop and demonstrate practical guidance against which dispersion and ventilation of naturally ventilated modules can be designed and assessed.