Testing The Plastic
Research shows that not all composite intermediate bulk containers (CIBCs) are created equal when it comes to fire protection.
NFPA Journal®, January/February 2009
By Joseph Scheffey, P.E.
Shippers and manufacturers are always looking for easier, more efficient ways to move product. More than two decades ago, the use of intermediate bulk containers (IBCs), or “totes,” changed the materials handling industry. IBCs, which can store 25 percent more liquid in essentially the same storage footprint as a pallet load of four 55-gallon (208-liter) drums, are reusable portable containers commonly used to ship chemicals in place of metal, composite, or plastic drums. As the name suggests, they are of intermediate size, larger than drums but not as large as storage tanks.
IBCs can be made of many different materials, including a composite of plastic and steel. Composite IBCs (CIBCs) are combination containers constructed so that the inner tank that contains the liquid is encased in a separate outer support structure during transportation, storage, and use. The inner vessel is a liquid-tight container frequently made with a high-molecular-weight, high-density polyethylene (HDPE) to a nominal thickness on the order of 0.06 to 0.12 inches (1.5 to 3.0 millimeters) or 100 mils. The outer structure may be made of many different materials, including steel, plastic, fiberboard, or wood. CIBCs used to store hazardous materials must be able to withstand federally and internationally mandated hydrostatic, leakproofness, stacking, and drop tests. Because plastic reduces storage costs, is easy to handle, provides product quality assurance, and resists corrosion, CIBCs often considered the container of choice.
Among CIBC users are producers of hazardous liquids. There is growing concern today among regulators and insurers that hazardous liquid storage in plastic containers is anathema to good fire protection. They fear that multiple, cascading failures of many units of flammable and combustible liquids might easily overwhelm fire protection systems, destroying the storage warehouse.
In 2006, the Fire Protection Research Foundation (FPRF) undertook a study to determine whether the way CIBCs are currently listed and labeled and the protection criteria specified would help prevent such an incident.
Fire testing CIBCs
When the industry first started looking at fire testing CIBCs, it was unsure about the appropriate performance criteria. In response to manufacturer, insurer, and enforcer concerns about protecting these large containers from fire, the NFPA Technical Committee on Warehouse and Container Storage created Annex E for NFPA 30, Flammable and Combustible Liquids Code. Annex E was a suggested test protocol for developing protection for containers storing liquids. This protocol, which can be used by any industry to plan and test protection criteria, was first adopted in the 1996 edition of NFPA 30.
Between 1996 and 2000, the CIBC industry teamed with the FPRF to conduct a series of fire tests. Failure mechanisms were scoped, palletized and rack arrays evaluated, protection criteria identified, and standardized test methods developed.
An important aspect of the FPRF program was the development of a listing program that allows manufacturers and third-party testing and listing agencies to test and approve individual plastic or composite IBCs for potential storage in warehouses. As part of the testing, a reduced-scale fire test method was developed and scaled to the results of the large-array testing, giving manufacturers two different methods of testing their IBCs. The resulting test method is UL 2368, Fire Exposure Testing of Intermediate Bulk Containers for Flammable and Combustible Liquids.
As a result of these large-scale tests and the promulgation of the UL test method, the 2000 edition of NFPA 30 adopted IBC palletized and rack storage protection criteria tables for Class II and Class III liquids, which allow unlimited storage of CIBCs when all the requirements of protected storage are met. If liquid storage in a building is to be considered protected, the mandatory protection criteria found in Chapter 16 of the 2008 edition of NFPA 30 must be used. These criteria are specific for the container type, class of liquid stored, and storage configuration. Additional parameters include storage and building height, as well as water or foam fire suppression system application rates required to control real-world fires. If these criteria are not met, a building is considered unprotected, and most of the contents of the fire compartment of origin will probably be lost should a fire occur.
The definition of CIBCs was also added (see sidebar). The units must be listed and labeled by UL — that is, classified, in accordance with UL 2368.
Further tests to address lingering issues
Almost a decade after NFPA 30 recognized CIBCs, they are relatively widely used to store both combustible and flammable liquids, but only one manufacturer currently produces UL-classified CIBCs. Why?
Lack of enforcement by code officials and lack of recognition of UL-classified CIBCs by the insurance industry are often cited. There is also a perception that there is not much difference between a listed container and a non-listed container, except cost. It is also difficult for a fire inspector to tell the difference between the two, although their fire performance is significantly different. Only by carefully inspecting the exterior markings can one see the classification markings on a fire-tested unit.
Responding to a series of large-loss fires involving CIBCs, The United Kingdom Health and Safety Laboratories fire tested CIBCs filled with alcohols and high-flash-point diesel fuel in 2005. (The results appear in “Fire Performance of Composite IBCs,” by G. Atkinson, Health and Safety Laboratory Report FR/05/09, Derbyshire, United Kingdom, October 30, 2005.) Testing concluded that a fire in a high-flash-point liquid may result in quicker container failure than one in lower-flash-point liquids, and that fuel-filled containers may fail more rapidly than water-filled containers, which raised additional concerns in the insurance industry.
The FPRF answered the call for data, organizing a group of interested parties to address these concerns and perceptions in 2005. The program was conducted in two phases.
Phase 1: Combustible liquid testing
The Phase 1 evaluation involved three palletized storage array tests conducted at the Southwest Research Laboratory in San Antonio, Texas. The objective was to determine whether non-listed CIBCs storing combustible liquids could be protected. Researchers chose a warehouse storage scenario 30 feet (9 meters) high. CIBCs filled with mineral seal oil, a Class IIIB liquid, were stacked two wide by two long by two high. The exposure fire was a 14-by-14-foot (4-by-4-meter) square pan containing mineral seal oil with a flash point of 270ºF (132oC).
The 275-gallon (1,041-liter) CIBCs used had an inner HDPE bottle wrapped in an outer layer of galvanized steel. The only area without galvanized steel covering was the fill cap area in the top.
Test 1 was performed with 286oF (141oC) standard response sprinklers discharging at 0.60 gallons per minute/square foot (2 liters per minute/square meter). Several CIBCs failed about 3 minutes after exposure to a sustained flue fire.
Test 2 was designed to provide faster sprinkler operation and improved suppression of the flue fire using aqueous film-forming foam (AFFF) discharged through 155oF (68oC) standard response sprinklers. The fire was extinguished approximately 2 minutes after the first sprinkler activated, with one CIBC suffering small holes and no substantial oil loss.
Test 3 was proposed to repeat Test 2 with 286oF (141oC) sprinkler heads, but the results were inconclusive due to test set-up problems.
Phase 2: Combustible liquid testing
The objective of Phase 2 was to address three questions that persisted following Phase 1 testing. Would listed CIBCs provide improved fire resistance in identical fire situations? Did pan tests represent a more severe fire threat to containers than the open, warehouse-type fires used in the CIBC listing process? And would CIBCs filled with combustible liquids have different fire-resistance performance than those filled with water?
Test 1 of Phase 2 repeated the pan-fire scenario using listed IBCs instead of non-listed units. A number of IBCs developed distinguishable leaks during the fire, and a post-test inspection revealed that five failed catastrophically. While the failure times could not be exactly determined, there was no distinguishable difference between the results of this test and the results of Test 1 of Phase 1. This indicated that the pool fire exposure is more severe than the floor fire exposure used in previous FPRF tests and adopted as an alternative evaluation method in UL 2368.
Test 2 was conducted in accordance with the UL 2368 large-array test method using UL-classified IBCs filled with water on which floated heptane. Forty-three seconds after ignition, two sprinklers operated, the only two to operate during the test. Results of this test were dramatically different from those of Test 1. Post-test observation indicated three damaged IBCs. One had a slight slit leak that resulted in weeping of the stored liquid. Another had holing in the container top above the liquid level. And the fill cap of a third had burned off.
The third test was a repeat of Test 2 using mineral-oil-filled, UL-classified IBCs. The objective was to see whether the oil had any effect on failure mechanisms or failure times. The first sprinkler operated 27 seconds after ignition. A total of four sprinklers operated, three of which operated in the 28- to 38-second range, and the array maintained its structural integrity.
Three units were damaged. A lower unit melted down and burned through at the top, above the liquid level. Another unit leaked at the valve and around the discharge valve. A third unit had a small leak at the face that resulted when the unit bulged from the heat, causing the internal ceramic insulating paper within the CIBC to pull apart at the seam. A stress crack resulted when bare plastic was exposed to the hot outer shell.
The CIBCs in Tests 2 and 3 failed to pass the UL leakage criteria, which require the prevention of liquid loss below the CIBC full liquid level. The listed CIBCs supplied for these tests apparently differed in construction details from the units originally tested, which probably contributed to the breaches. Researchers concluded that sections of the internal ceramic insulating paper should overlap so that they would not separate and expose bare plastic to the outer metal container.
Tests 2 and 3 essentially achieved all other NFPA 30 performance criteria. Researchers saw no significant differences in the performance of the water- and oil-filled containers; they surmised that the insulating paper in the listed CIBCs reduced the temperature the plastic bottles reached during testing. Bottle temperature appears to be a major variable affecting the rapid failure rates involving the oil-filled HDPE bottles in the UK tests.
NFPA 30 is a mandatory flammable and combustible liquid regulation in many U.S. jurisdictions. Occupancies storing or using CIBCs are subject to significant limitations of unprotected storage, unless:
- The CIBCs are listed and labeled in accordance with UL 2368 or an equivalent test procedure,
- Approved storage configurations are used,
- Adequate fire protection, particularly high-density ceiling and rack fire suppression system capacity, is provided, and
- Storage is limited to Class II and III liquids.
A decade of testing has demonstrated the reaction of CIBCs to large fire threats. Researchers have developed alternative test methods and quantified their differences, and have shown the differences between individual CIBC designs, which is the fundamental reason for adopting a storage fire test and listing requirement. CIBCs’ need for some degree of inherent fire resistivity when used to store combustible liquids is clear. Research has also demonstrated the effectiveness of protection options, using either water or AFFF. The most recent CICB test reports can be downloaded from the FPRF site.
Continued fire testing and field experience may change the way large containers are listed and labeled and the protection criteria specified. Any fire professional dealing with the handling and storage of liquids should stay informed on this important topic.
Joseph L. Scheffey, director of RDT&E at Hughes Associates, Inc., is a member of the NFPA 30 Technical Committee on Storage and Warehousing of Containers and Portable Tanks. Contributing to this article were Michael Snyder of the Dow Corning Corporation; Christina Francis of the Procter and Gamble Company; Robert Benedetti, NFPA liaison to the NFPA 30 Technical Committee; and Kathleen Almand, executive director of the FPRF.