Fire resistance of passive fire protection coatings after long-term weathering

https://doi.org/10.1016/j.psep.2009.09.003Get rights and content

Abstract

Passive fire protection (PFP) systems are widely used by the oil, gas and chemical industries to protect offshore and onshore facilities against the effects of fire. However, there are concerns that the performance of PFP systems in a fire may have deteriorated because of weathering and/or that corrosion of the protected item may be taking place beneath the PFP systems. In May 1987, Shell UK began a long-term PFP weathering programme at a maritime test site using furnace tests to assess the fire performance of the materials. The programme was handed over to the Health and Safety Executive in October 1999 and was continued and extended to include the more severe jet fire testing of weathered specimens. This paper describes the effects of weathering on six epoxy intumescent PFP products and one cementitious PFP product in common use. The results are discussed in relation to alterations in physical form, corrosion of the substrate and changes in fire resistance.

Introduction

Passive fire protection (PFP) coatings are widely used in the offshore industry and increasingly so for onshore applications. They are relied upon to provide front-line protection to structural steel, walls and sometimes vessels and piping or valves, i.e. to safety critical components, in the unlikely event of a major fire and must therefore retain their fire resistance performance throughout the lifetime of the facility. In the early eighties, Shell was concerned that no guarantees could be given that the PFP being used would still be fit for purpose a decade or more later. Of particular concern were the quite new (at that time) intumescent coatings that were increasingly being specified. After years of exposure to the harsh North Sea environment, it was possible that the active ingredients, essential to intumescence and the formation of the protective char layer, could be leached out. A survey of the PFP installed on a number of Shell North Sea platforms was carried out in order to identify the location and to assess the condition of materials then in service. The survey revealed that whilst some were apparently in good condition, others were obviously not. Documentation was not always available, rendering identification difficult. Overall, the survey reinforced the need to generate information on long-term durability for maintenance purposes and to aid future selection.

Three approaches were possible, viz.:

  • (1)

    removing and fire resistance testing of existing sections of material,

  • (2)

    accelerated weathering using climatic chambers, or,

  • (3)

    development of a programme of real-time weathering under known exposure conditions.

The first option did not allow prediction of what would happen to the new materials and, in many cases, insufficient documentation was available to adequately identify the PFP used. Accelerated weathering is widely used for paint systems and is used for PFP (UL 1709, 1998, NORSOK M 501, 2004) but the climatic chambers available at the time were of insufficient size to weather a reasonable number of the relatively large size specimens needed for fire resistance testing. As the above standards did not exist in their current form and had not been validated by weathering PFP, it was necessary to perform some real-time weathering in order to determine the acceleration factor for PFP materials. Shell therefore decided to begin its own programme of weathering samples of PFP, suitable for subsequent fire testing, at a maritime site.

The aims of the original programme were to:

  • (a)

    contemporaneously test all of the commonly used products at a single facility,

  • (b)

    expose the samples to a demonstrably similar environment to that experienced in the North Sea,

  • (c)

    inspect the samples regularly and note and photograph any deterioration,

  • (d)

    remove samples periodically and subject them to standard furnace tests (BS 476 [BSI, 1987]) to determine any long-term variation in the fire resistance of the PFP coatings as measured against their ‘as new’ performance.

The first phase of the Shell programme was started in 1987, using specimens designed to give 60 min protection against a cellulosic fire, and was terminated in 1997. Two basic types of coating were examined; three intumescent (char-forming) products and three products based on cementitious materials. Two of the intumescents, Chartek III and Pitt-Char, survived unaffected throughout most of the 10 years. There were only minor changes in appearance. The other intumescent product, Firec 123 was not provided with a topcoat and it suffered from erosion. The mass of coating lost exceeded any mass gain due to water absorption. Of the cementitious products, Pyrocrete 201, which is based on magnesium oxychloride, was found to promote corrosion of the substrate and Shell's approval for its own use offshore was withdrawn at an early stage. Another cementitious product, Contraflam 3810, was found to absorb unacceptably large amounts of water. A caution concerning its use was issued by Shell and it was eventually removed from the market. The remaining cementitious product, Mandolite 550, relied heavily on protection from its topcoat to avoid massive erosion. This product also absorbed significant quantities of water, which augmented its fire protection capability. Overall, it was concluded that the topcoat had an important role in determining the longevity of the bulk fire protective coating, irrespective of type. Failure of the topcoat by erosion or cracking leads to ultra-violet (UV) attack of the bulk coating in the case of intumescents and erosion and water ingress into cementitious materials. Chartek III and Pitt-Char physically survived the Phase 1 weathering programme, did not exhibit extensive corrosion and when subjected to a fire test, gave results that were either equivalent to those from the virgin materials or a credible variation, e.g. a slower temperature rise due to absorbed water. In all aspects, these intumescent products were far less affected by weathering than were the cementitious products.

In 1991, a second phase was started using specimens designed to give 120 min protection against a hydrocarbon pool fire, as demonstrated by furnace testing (BS 476 Part 20 Appendix D). The testing showed no real fall off in fire resistance and hence Shell intended to finish Phase 2 of their programme in 1998. In the report (The Hon. Lord Cullen, 1990) on the public enquiry into the Piper Alpha disaster, it was recommended that work was needed to address the risks of hidden corrosion occurring beneath PFP coating systems and their performance in jet fire conditions. The United Kingdom (UK) Health and Safety Executive (HSE) responded by setting up a research and development programme dealing with these areas of concern. In particular, HSE (in conjunction with the then Norwegian Petroleum Directorate) took the lead in developing a standardised, intermediate scale test (Jet Fire Test Working Group, 1995) to assess the resistance of PFP coatings to jet fires. Subsequently, HSE's Health and Safety Laboratory (HSL) was commissioned to assess whether this test could be modified to allow comparative testing of the specimens being weathered by Shell. A procedure was successfully developed and hence HSE was in a position to adopt Phase 2 of the Shell programme, take it to the 10-year stage (longer for some products) and supplement the programme with an assessment of the jet fire resistance using a standardised test (Jet Fire Test Working Group, 1995, now ISO 22899, 2007). This was more likely to show up any deterioration in fire resistance due to weathering. As this was an extension to the original work, it is noted that none of the manufacturer's panels were correctly specified for jet fire resistance – this is discussed later.

HSL managed the extension to the project and performed the jet fire resistance testing. Shell provided the data generated up to 1998 and continued to provide technical advice. CAPCIS was retained to operate the facility and monitor the samples for deterioration, including corrosion. Warrington Fire Research continued to perform the furnace tests. This paper describes the results from the extended Phase 2 programme.

The effects of physical damage (gouges, nicks, disbondment, etc.) on fire resistance and the adequacy of repairs are not addressed in this paper. However, a joint industry project has been recently completed which investigated the effects of different types of damage, deliberately induced in new material, on jet fire resistance. For further information on this and the effectiveness of different types of repair, see Kerr et al. (2008).

Section snippets

Product description and specimen preparation

It was decided at an early stage that the definitive monitoring tool should be a fire resistance test, because this ultimately reflects the primary function of PFP materials. The fire resistance of PFP cannot be determined in isolation as the PFP must be formed upon and bonded to a substrate (usually steel) representative of that which is to be protected (note: neither cast nor composite systems are considered in this paper). For BSI 476 Part 20 fire resistance tests, a stiffened steel bulkhead

Weathering

The basic requirements for real-time weathering are direct and continuous exposure to the natural environment. Ideally, exposure on an offshore platform would have provided the most credible conditions. However, there were factors other than the general environment to consider, for example:

  • -

    unobstructed space to mount specimens so that they are fully exposed and not overshadowed,

  • -

    unrestricted access for monitoring as and when the environmental conditions, state of the specimens, or the

Corrosion assessment

Coating breakdown, that is of the total system as compared to the effects of weathering evident in the topcoat, and corrosion of the substrate was not discernible due to the overall thickness of the coating systems. However, when corrosion occurred, it was generally evident at the edge of each panel. In order to determine whether corrosion had occurred beneath the PFP coating systems, the coating was removed, where possible, from each panel using a hammer and cold chisel. Some corrosion

Furnace testing

As indicated earlier, the fire resistance test used by Shell was a variant of BSI 476 Part 20 (BSI, 1987). The Appendix A heating regime was used for “cellulosic” fire tests (slow heat up time and furnace temperature ca. 945 °C after 60 min at the end of the heat exposure time) and the Appendix D regime for “hydrocarbon pool” fire tests (faster heat up time and furnace temperature ca. 1100 °C after 30 min with a total heat exposure time of 120 min). A frame was designed for a wall furnace to take 16

Jet fire testing

The realisation that some PFP may have to resist high-pressure jet fires have led to the development of the Jet Fire Resistance Test (Jet Fire Test Working Group, 1995), now ISO 22899 (2007). Whilst the coatings used in the Phase 2 programme were not intended to protect against such a fire scenario, some of those being weathered have been found to perform satisfactorily, albeit with a slightly thicker coating. However, it was known at the outset that the Pitt-Char specimens would not perform

Coal tar epoxy masking coat

A coal tar epoxy coat was applied to mask the back and edges of the test plates to prevent (minimise) corrosion of the steel substrate and to provide an additional barrier layer to mitigate edge effects that would not be experienced for coatings applied to a larger structure. It was subsequently lost at many areas along the edges and where it overlapped the topcoat at the front face. Shell inspection records noted the occurrence of “environmental stress cracking (esc)” of the topcoat for

Chartek III

The Chartek III had a rough textured surface with a degree of variability in the finish. No topcoat had been provided and the bulk coating had a slightly yellow, off-white colouration. The weathering effects observed for the panels are summarised in Table 3. The detail of the surface of one panel (showing surface cracking, staining and delaminating) at the end of the exposure programme is illustrated in Fig. 4.

The surface roughness of the Chartek III had increased markedly upon weathering. The

Chartek IV

Unlike the Chartek III specimens, the Chartek IV specimens were supplied for exposure with a white, high-gloss topcoat. The weathering effects observed for the panels are summarised in Table 5. In Fig. 6(a), one of the panels remaining is shown at the end of the exposure programme before retrieval. There was little sign of cracking or staining.

Examination of the substrate after removal of the PFP material in the grey areas shown in Fig. 6(b) indicated that the steel panel appeared to have been

Firetex M90

The bulk coating had a distinctive light blue colouration, from a pigment added to one of the components as a quality control aid, which showed through where the topcoat was damaged. For delivery to Shell, the specimens had been stacked in a box with no packing between them. In transit, the protruding ends of the fixing inserts on each of a number of specimens had damaged the topcoat of the underlying specimen. These constituted the main features logged in the initial inspection following

Mandolite 550

A number of pinholes and micro cracks in the topcoat were noted before exposure with discrete areas of topcoat missing. The cementitious material had high water content (26.0% of dry material). The topcoat failed after 2 years exposure and the panels were returned to Mandoval. Existing topcoats were scraped off and loose material vacuumed off. The panels were then kept at 20 °C for 6 days with significant weight (water) loss, although a steady state was not achieved. The replacement topcoat

Nullifire system E (Mandolite 990)

Nullifire System E (Mandolite 990) was supplied for weathering with a white, high-gloss topcoat. The only features other than on the deliberately damaged specimen were some small stained areas where globules of the coal tar epoxy edge sealant had been broken off. The weathering effects observed for the panels are summarised in Table 11. A panel before removal is shown in Fig. 13(a).

Corrosion was apparent (see Fig. 13(b)), principally at the bottom two corners and to a lesser extent the top left

Pitt-char

The topcoat of Pitt-Char was smooth and glossy but the texture of individual specimens varied considerably. There was some loss of topcoat from high spots prior to exposure but occurrences were few and scattered. The bulk material was particularly tough and resistant to cutting. Shell observed that there appeared to be a rubbery primer between the steel and the main coating layer. The bulk coating was anchored to the steel by five individual pins with large diameter (30 mm), spring washers

Thermo-lag 440

No information concerning the preparation of the specimens was provided. The major face of the Thermo-lag 440 specimens had a smooth, uniform gloss white finish with few flaws. A section cut through a specimen revealed a grey primer on the steel and a squared wire retaining mesh located about 10 mm above the steel. Extensive corrosion of the rear and edges of the steel plates caused extensive rust staining of the white topcoated faces (see Fig. 20(a)). Localised areas of cracking in the topcoat

Conclusions

One cementitious product (Mandolite 550) and six epoxy intumescent products (Chartek III and IV, Firetex M90, Nullifire System E, Pitt-Char and Thermo-lag 440) have been subjected to long-term weathering at a maritime site. The conclusions in relation to their resistance to weathering, corrosion and fire are as follows.

Weathering

The topcoats applied to the epoxy intumescent PFPs in this programme appear to have performed satisfactorily for up to 10 years before beginning to exhibit general breakdown. Localised breakdown was observed sooner at various surface features such as high spots, included fibres, etc., i.e. regions of higher stress. Topcoat breakdown was apparently related to surface roughness. The intumescent PFPs generally have a rough surface. Normally, the acceptable surface finish is agreed between the

Corrosion

No corrosion was observed for the specimens of Chartek III (10.5 years), Chartek IV (8.5 years) and Firetex M90 (7.5 years). Corrosion occurred on the Thermo-lag 440 (10.5 years) and Pitt-Char (10.5 years) and Nullifire System E (8.5 years) specimens that had initiated at the edges and progressed inwards. The differences in the extent of corrosion that has occurred beneath each coating system may be explained by the apparent quality of the test panel preparation and protection afforded to the

Fire resistance

Conclusions relating to any changes in fire resistance are summarised in Table 16.

The improved fire resistance of Mandolite 550 was due to absorption of water. This may also be true of the Thermo-lag 440. The slight reduction in jet fire resistance of Firetex M90 is attributed to it being the only product (apart from Pitt-Char, which had no mesh reinforcement at all) not to have wire mesh reinforcement (glass fibre mesh was used instead). The Pitt-Char showed no reduction in fire resistance in

Overall

The project has revealed a reassuring picture overall. In particular, the products in most common use have shown either no or only slight deterioration in fire resistance over 10 years. There were no corrosion problems beneath three of the epoxy intumescent coatings and the corrosion associated with the other three epoxy intumescent coatings clearly originated at the edges. There were considerable corrosion problems with the cementitious product but it was still fully fire resistant as the

Acknowledgements

The authors gratefully acknowledge the support of Shell, the Health & Safety Executive and the PFP manufacturers in this work, also the contribution of the personnel performing the considerable number of tests involved: D. Willoughby (HSL) and D. Forshow (Warrington Fire Research).

The authors especially acknowledge the enormous contribution made by the late Dr. Terry Cotgreave (Shell) who conceived the original Shell programme and worked meticulously on it over almost two decades.

References (8)

  • British Standards Institute (BSI), 1987, BS 476. Part 20: British Standard Fire tests on building materials and...
  • International Standards Organisation (ISO), 2007, ISO 22899-1, Determination of the resistance to jet fires of passive...
  • International Standards Organisation (ISO), 2007, ISO 8501-1, Preparation of steel substrates before application of...
  • Jet Fire Test Working Group, 1995, The Jet-Fire Resistance of Passive Fire Protection Materials, HSE report OTI 95 634,...
There are more references available in the full text version of this article.

Cited by (0)

View full text