This information sheet provides guidance on water deluge systems, their testing requirements, data recording in relation to performance standards, and new knowledge of fire attack scenarios. It is noted that deluge testing offshore is generally being carried out as required. However some flow measurements are not being recorded effectively, and data recorded is not being analysed to provide the best information for use in optimising maintenance and performance. This note does not cover water mist, or spray systems in enclosed spaces.
In the event of fire on offshore installations, the primary protection of pressurised process vessels and pipework is a combination of emergency depressurisation (EDP) via the blow-down system and passive and/or active fire protection. The EDP systems are currently designed to industry guidance American Petroleum Institute Recommended Practise (API RP 521)1. However, it is widely recognised that in some respects this is inadequate and inappropriate for offshore installations, for which it was not originally intended2. For example, in designing an EDP system only heat transfer from ~100 kW m2 fires to the liquid wetted wall is considered. A major concern is that heat loads from fire attack, quoted in the current guidance, are much lower than can be expected. Based on latest knowledge2 severe fires that may occur on offshore installations may generate thermal fluxes of up to 400 kW.m2 and thus EDP may not guarantee vessel or structural steel protection.
In a fire attack situation steels’ strength starts to decrease significantly once a critical temperature is reached; generally taken as ~450-500oC. With a protected vessel (passive or active protection), EDP reduces the internal applied stress by removing the internal vapour. If the same vessel does not have fire protection the higher heat flux on the vessel walls causes the vessel temperature to rise more rapidly and hence the vessel wall strength to decrease more rapidly. The vessel may be depressurised at the same rate as before, but the wall strength is dropping too rapidly and the vessel will fail (rupture) once its strength is less than the applied internal stress caused by the internal vapour pressure.
Structural steel is usually protected from fire by passive fire protection but in some cases also by water sprays. EDP can benefit structural steel work under fire attack by reducing the time of jet fire impingement.
ISO standard, ISO 13702:1999 (1-4)3, has set out the requirements and guidelines for the control and mitigation of fires and explosions on offshore production installations.
A water deluge system can either be designed for general area coverage, or a directed water spray protecting a specific equipment item. Both are delivered from fixed pipework and independent fire pumps and provide protection to:-
Historically most active fire protection systems offshore were designed for compliance with SI 611, "Offshore Installations: Guidance on Fire Fighting Equipment". Current legislation is goal setting but SI 611 remains a good basis for reference. Other standards containing useful information include ISO 104184 and API RP 14C5.
OTO 2000 0512 provides useful details of fire types, thermal fluxes, source terms, flame geometries, combustion products that may be useful in an installations fire and explosion strategy (FES) analysis. Table 2 for example, provides further guidance on the uncertainties in the knowledge of certain fire types. It also references the original HSE OTI Reports that may have been used in the original safety cases. This may be beneficial to newer duty holders if they have difficulties in finding the origin of the datasets used in the initial hazard analysis of their acquired installations.
The effectiveness of water deluge systems is significantly enhanced when used in conjunction with a fire retarding foam. Aqueous film forming foam (AFFF) when added to water spray systems forms a foam blanket over a pool of burning liquid and is highly effective in extinguishing a pool fire. Fire extinguishing times of foam blankets are significantly faster than those of water deluge alone. Foam blankets are most easily formed and retained when the liquid spill is contained in a bund allowing a rapid foam buildup within the confines of the bund. This has the added advantage that the burning hydrocarbon fire does not move around the deck. Please see Offshore Information Sheet 9/2008 - Modeling pool fires in offshore hazard assessment6 for additional guidance.
It is recommended that all main process vessels should have a liquid retaining bund beneath, connected to the hazardous area drain system, via a drain valve that removes liquid from the bottom of the liquid pool. Removing liquid from the pool bottom preserves the fire extinguishing properties of the foam blanket by not disturbing its surface. The drainage system should be designed to cope with the full rate of discharge, both the deluge water and the vessels inventory.
The prime objective of a water deluge system is to cool equipment in a fire incident. Cooling water should arrive onto a designated target (valve, vessels, critical structural component etc) as soon as possible, to maintain the targets’ temperature below 100oC. Any surface above 100oC will be difficult to cool as the water will tend to boil off and the surface will not “wet”. If a surface is significantly above 100oC, it may become very difficult to wet and hence a hot spot will occur. In process vessels this has the potential to cause asymmetric stress in the vessel wall, leading to possible catastrophic failure in a very short time span i.e. minutes.
ISO 13702 provides generally accepted specifications for a water deluge system for general area coverage (Table C.3) some of which is reproduced below:-
A general area coverage of 10 – 12 Litres min-1 m-2 (LPMM) to protect most platform area against pool fires, with 20 LPMM recommended for pumps/compressors and 60 LPMM for chainlockers.
It is advised that these are minimum flowrates and that larger flowrates and accurately aligned nozzles can only benefit the fire protection of both vessels and structural steel.
Specific or directed deluge is sometimes employed instead of, or as well as, the general area protection. Individual items, such as vessels or heat exchangers, are protected as described in National Fire Protection Association standard NFPA 157. Most offshore water deluge systems provide flowrates in excess of these values. For wellhead valve systems a rate of 400 litres per Wellhead is recommended.
Nozzle alignment to direct water accurately onto its target is recognised as a key parameter in deluge effectiveness. The number of blocked nozzles is also significant in that uneven distribution of water may result. During routine maintenance and performance evaluation of the number of blocked nozzles, and the pattern of blockage, should be measured and recorded. A performance standard for the maximum number of blocked nozzles should be based on the FES analysis of each fire area. Additionally, the blockage pattern should also be observed and recorded. Random distribution of blockage is acceptable; a cluster of blocked nozzles is not, as it could result in an uneven distribution of cooling.
Tests carried out by HSL and Shell Global Solutions2 on deluges and liquefied petroleum gas (LPG) tanks indicated that a considerable amount of the water directed at the tank may miss the target. Hence, an application rate calculated as the amount of water leaving the nozzles divided by the tank surface area is a poor measure of how well a tank is protected by a deluge system. It should be noted that water deluge systems are generally designed with about a 30% overcapacity to account for wind shear.
It is important that during a water deluge test the water flowrate at the target is measured and not just the water flowrate leaving the nozzles. i.e. elephant trunking can give a false, optimistic, measure of water flowrates. Accurate water flow onto a target is best achieved by the use of 2 or 3 trays laid around the local area that will provide a more accurate measure of the water volumes averaged from several nozzles. If a tray of 1m2 area is used water flowrate measurement is a straightforward calculation.
Fig 1.1 - Vessel survives
Fig 1.2 - Vessel fails
The analysis of each fire scenario on an Installation should focus on the two illustrative graphs (Figs 1.1 & 1.2) contained on page 1 (Introduction) of OTO 2000 051 (see below). Control and protection systems should be developed such that the stress and time curves are in line with Fig 1.1 particularly for hydrocarbon containing vessels. Fig 1.2 illustrates failure mode from fire attack.
To generate the best and useable information from deluge test data a simple test data table is recommended. An example is given below with arbitrary performance standard values included. Accurate performance standards need to be based on the fire analysis of each designated fire area of an Installation. OTO 2000 051 probably contains the best current advice for developing scenario specific fire attack and protection strategies.
| Fire area.
“Process Area 1”
|Performance Standard||Previous Measurement||Current Measurement
|Duty Pump tested||Standby Pump tested|
|Time of water flow from most remote nozzle (s)||<30 seconds||15secs||18secs||X|
|Pressure at control valve (bar)||5.5barg||7.2barg||6.4barg|
|Response time for control valve to operate||5secs||3.5secs||4.3secs|
|Visual Inspection of nozzles –No. blocked||5% evenly distributed||3.8%||4.5%|
|Number laydown trays used||3||3||2|
|Flowrate onto trays||12 litres/m2/min||14.5||15.0|
|Adequate drainage?||Visual confirmation||Yes||Slow drain away|
|Water pressure at Fire pump||14.5 barg||14.5||13.5|
The performance standard column should be populated with data from the original hydraulic balance and the fire hazard analysis of each designated fire area or zone. This is usually found in the installations safety case. Account should be taken of any changes in the platforms hazard profile in terms of increased or reduced fire hazard. It is advised that the installations hazard profile and associated fire hazard analysis is reviewed periodically but not less than 5-yearly.
A performance standard for nozzle blockage should be based on the fire hazard analysis of each fire area. It is generally accepted in the UK offshore industry that 5% nozzle blockage is a failure criterion and that several blocked nozzles occurring on one leg of an area deluge is also regarded as a failed test.
It is recommended that the flowrates are taken from the volumes captured in the laydown trays and not from ‘elephant trunking’ around the nozzles. A simple isometric drawing of the nozzle arrangements for each fire area could be used to mark any blocked nozzles, to provide a quick effective guide to systematic blockage.
Offshore Installations (Safety Case) Regulations 2005 (SCR05) – Regulation 12(1) requires that all major accident risks have been evaulated and measures have been, or will be, taken to control those risks to ensure that the relevant statutory provisions will be complied with.
Offshore Installations (Prevention of Fire, Explosion and Emergency Response) Regulations 1995 (PFEER) – Regulation 13(a) & (b) Mitigation of fire and explosion, Regulation 14(1) Muster areas etc and Regulation 19(1) Suitability and condition of plant
This information sheet contains notes on good practice which are not compulsory but which you may find helpful in considering what you need to do