Keeping Older Oil and Gas Facilities Safe
Assessments of one U.S. and two Latin American refineries show where and how to improve fire protection
by Roberto Clark, P.E., and Fernando Escalante, P.E.
The petrochemical industry is one of the most investment- and capital-intensive markets in the world today, with a tremendous capacity to affect both national and international economies in the event of a catastrophic accident. It's also one of the most hazardous industries in the world in terms of product flammability, volume of processes, and extreme operating conditions.
Given the fact that most refineries in the United States and Latin America are more than 25 years old, many still with their original fire protection systems, the owners and insurers of major refineries worldwide are increasingly concerned about fire and explosion safety at today's oil and gas refineries.
The Project Refineries
Two of the world's largest petrochemical companies, both headquartered in Latin America, and one medium-size refining company in the United States are currently analyzing and upgrading the fire-protection systems of several of their most critical refineries. Three of them, each of which produces between 160,000 and 240,000 barrels of distillates, gasoline, bunker, kerosene, and diesel a day, are 28, 34, and 30 years old. Their product is stored in 200 oil and gas products storage tanks that each hold up to 1.386 million gallons (5.246 million liters or 330,000 barrels). Two refineries, one in Brazil and the other in Venezuela, are among the biggest in South America, and the other in Ohio is a medium-size refinery.
Fire brigades currently protect all three facilities. Their fixed protection systems include 2 to 4 miles (3.2 to 4.6 kilometers) of fire mains fed by numerous fire pump stations, dedicated water tank or reservoirs, yard hydrants and fixed monitors, tank foam systems, and water-spray systems. A European firm originally designed the Brazilian refinery and its fire-protection systems, while American firms designed the other two refineries.
As part of the petrochemical companies' update program, the consulting firm of Rolf Jensen & Associates was asked to analyze the refineries' water supply, active and passive fire-protection systems, risk exposures, fire brigades, and emergency plans. As expected, we found them both to have the sound, strong housekeeping practices and solid, well-trained fire brigades for which oil and gas facilities are known. In fact, personnel in their fire brigades were trained at the Texas A&M Firefighting School. However, our systems analysis produced some notable and interesting findings.
The initial step of our analysis was to develop and establish a set of test protocols to assess system conditions. We based the testing protocols on the commissioning and acceptance test procedures required for new systems, in accordance with NFPA codes and standards. Among those we used were NFPA 20, Installation of Stationary Pumps for Fire Protection; NFPA 11, Low-Expansion Foam; NFPA 13, Installation of Sprinkler Systems; NFPA 15, Water Spray Fixed Systems for Fire Protection; NFPA 24, Installation of Private Fire Service Mains and Their Appurtenances; and NFPA 72, National Fire Alarm Code®.
Using the systems' testing results, we conducted a detailed investigation into the water sources and the pumping and supply systems. This phase of the project included developing actual fire-pump characteristic curves, fire-main rugosity conditions, and C-factor analysis and evaluation.
The fire-pump performance-curves indicated signs of premature damage and a lack of maintenance, probably because the fire-main water was used for process purposes, which may mean that the fire pumps were set at a fraction of the required fire-suppression system pressure to avoid damaging facility process connections. Using the fire-main and supply systems for facility process may not only cause premature corrosion in unlined fire-main pipes, but it may also jeopardize the refinery's overall safety level. In addition, one of the facilities' emergency procedures required that the fire pumps be re-set to higher pressures to feed the fire-protection systems properly, thus introducing the human-error factor into the mix.
One of the most important parts of our evaluation was the C-factor analysis. This type of analysis is related to the C parameter of the Hazen-Williams flow calculation used to calculate the friction loss; C-factors are related to the internal rugosity of piping. We used hydraulic models to estimate C-factors by comparing actual flow tests at hydrants with the simulated hydraulic behavior of piping.
We divided the fire main into small sections based on the type of pipe, the date of the original installation or replacement, its history of leaks, and similar water-flow demand scenarios. We collected flow and residual-pressure data from hydrants and monitors flowing at representative positions in each pipe segment of each fire-main quadrant or section. We then used this data, together with actual condition of the fire main, to generate a hydraulic model of the site, which we tuned by running hundreds of calculations so we could select C-factor values that aligned with the field-test results of water-flow tests.
Because there were the thousands of feet of fire-main piping and hundreds of hydraulic nodes in the fire-main model, the only way we could achieve realistic results was to go through the systematic, time-consuming process of trial and error. The estimated C-factor analysis of the fire main, combined with other testing, such as an ultrasound measurement of the pipe thickness and records of system leaks and replacement, provided a good estimate of internal condition of the fire-main pipe.
Low C-factor numbers indicated that, in spite of the fact that it was cement-lined, the fire main piping had poor hydraulic behavior. Further investigation and analysis revealed several hidden factors. A local company had lined the piping without following procedures approved by the American Water Works Association. The cement lining wasn't concentric, so it wasn't evenly distributed, leaving parts of the piping protected by very thin layers of cement that were susceptible to cracking and erosion. And the pipe was regular Schedule 40 pipe, which had been welded on site. This meant that the junctions and fittings were poorly protected or completely unprotected, which led to erosion and premature corrosion.
In another refinery, the regular Schedule 40 steel pipe, aboveground fire main was gradually and severely damaged by its use in cooling equipment and vessels in the process areas. In spite of the low flow, the permanent oxygen transportation into the fire main promoted internal corrosion of the pipe.
We used this information to plan for fire-main replacement or expansion, on which the petrochemical companies are spending, and will continue to spend, hundreds of thousands of dollars. The personnel managing the refineries' fire mains have found what sectors of piping need to be replaced and when to do so been a valuable decision tool.
The facilities' fixed fire-protection systems included water spray systems that protected the process area pressure vessels, foam-water sprinklers protecting the vehicle-fueling area, and semi-fixed foam systems that protected the storage tanks. Although these systems did operate, for the most part, they didn't comply with NFPA 11 in terms of discharge time and flow, systems concentration, nozzle or chamber selection, and spacing and location of devices.
Because there were no as-built drawings for these systems, we had to base our evaluation on field surveys, inspections, and water-flow tests. The findings were similar for two of the refineries: the water-spray systems weren't reliable, in some cases consisting of drilled pipes or small steel nipples instead of approved nozzles connected to a pipe ring. We found that the nozzles of other water-spray systems were directed only at the upper part of the vessel, thus relying on water rundown to protect the remaining portions. When we considered the wind effects, this approach didn't provide adequate protection. In addition, the severe lack of maintenance allowed nozzles to become plugged with insects, dust, and other external objects and water to flow without strainers.
The majority of the discharge chambers in the tank semi-fixed foam systems in one of the refineries seemed to have been manufactured on site, probably by the original tank provider. They didn't have the number of chambers stipulated in NFPA 11, jeopardizing the systems' overall reliability. In most cases, these chambers didn't have orifice plates, either, which created deficiencies in foam-solution proportioning and foam discharge time. Under testing conditions, we also found that the fire pumper truckers weren't calibrated properly to provide the proper water flow and pressures.
During these three projects, the owners asked us to determine whether the insurance company's requirement of providing fixed water-spray systems for exposure or radiation protection for the storage tanks was really necessary. Insurers frequently recommend such systems because they fear a tank fire will expose other tanks to heat and radiation risks, thus starting other fires or explosions. This concern isn't new. However, these systems have a significant price tag.
To answer this question, we performed exposure-driven engineering. The American Society of Chemical Engineers has identified a maximum acceptable radiant-flux-level value that can be used to determine the scenario, before hazardous damage can occur between storage tanks. We provided fire-modeling analysis that captured, on a tank-by-tank basis, the maximum expected radiant-heat-flux values, based on products, flammability, tank geometry, geographical location, and such weather conditions as temperature, wind, and humidity.
The results indicated that only three of the two refineries' 200 tanks had unacceptable radiant-heat-flux values, thus requiring a fixed water-spray system for radiation protection.
Improving the level of safety
These two cases can provide a universal approach to identifying the proper analytical and corrective tasks necessary to upgrade the safety levels of oil and gas facilities. Safety begins with a reliable water supply, the first line of defense for both facilities and their fire brigades. The majority of potential deficiencies arise in this area.
Despite the enormous efforts some of the largest refineries make in traditional risk analysis, the fire-protection systems' performance and design of refining facilities, often don't provide an acceptable level of protection. Although many owners have made an effort to modernize their fire-protection systems and implement the recommendations produced by fire risk analysis, recent fires in petroleum refineries are evidence that oil and gas facilities continue to burn.
The only way to determine whether your oil and gas facility is truly safe is to perform continual engineering analyses. Sound engineering, fire modeling, and loss-control basics will promote safer facilities.
Roberto Clark, P.E., is the engineering manager, and Fernando Escalante, P.E., a consulting engineer, both with Rolf Jensen & Associates, Latin America office.
|Model Establishes the Radiation and Exposure Protection Required for Refinery Storage Tanks
Model Establishes the Radiation and Exposure Protection Required for Refinery Storage Tanks
|Inside the Beltway
Inside the Beltway: New program to hire more firefighters dropped out of homeland security bill.
|Fuel-Cell Vehicles—Coming Sooner Than You Think
Fuel-Cell Vehicles—Coming Sooner Than You Think: Still, there are plenty of safety hurdles to master, including development of appropriate standards.
|Read how alternative fuels are becoming more common at filling stations
Read how alternative fuels are becoming more common at filling stations (PDF*, 991 KB).
|Bonding Versus Grounding
Bonding Versus Grounding: Why NFPA 407 no longer specifies grounding for safety during aircraft fueling.
|Inside the Beltway
Inside the Beltway: DOT looking at adopting NFPA 498 to protect shipments of explosives
|Inside the Beltway
Inside the Beltway: OSHA all but adopts NFPA 101® for its new exit routes standard (Members only**)
|No Early Warning
No Early Warning: German and French train fires illustrate need for alarms. (Members only**)
|Inside the Beltway
Inside the Beltway: Budget woes for federal forest firefighting portend long-term problems. (Members only**)
|Inside the Beltway
Inside the Beltway: NIOSH reviews comments on its draft standard for CBRN air-purifying respirators (Members only**)