Once again white water condensate clouds rise above the Texas A&M University fire grounds testifying that another LNG safety training and research event is about to take place.
BP continues to invest substantially with the Emergency Services Training Institute Fire Training School at Brayton Fire Training Field, College Station, Texas, to create a world class training and testing facility for the world to use. BP is using this to gain a full understanding of the best ways of detecting and handling LNG spills to minimize the hazards for their existing and proposed facilities worldwide.
Several of their front line emergency responder teams and mutual aid partners have already undergone intensive workshop training in realistic scenarios, using state of the art equipment, systems and technical knowledge. This has been provided by the Texas A&M specialist LNG team and a range of high performance participants who have been identified by BP to be world leaders in their fields.
The BP responders and their mutual aid colleagues have found the experiences exceptionally beneficial. These workshops have given them confidence and insight into how to safely handle LNG spills on land and water, to minimize their impact should an emergency incident happen. Let's not forget that LNG is already one of the safest industries that exists in the world today.
BP has been willing to share this information by allowing engineering firms who normally do not participate in these kinds of events to come as observers. At the last event AKER KVAERNER, currently involved in LNG, was present. KBR, Flour and JGC have all attended over the past year and a half to gain first hand knowledge in how LNG behaves so that front end engineering is on the same pages as first responders.
Co-sponsors of the BP LNG training project include Honeywell Zellweger, Angus Fire (Kidde) Micropack, Ansul, Detection and Measurement Systems, International Paints and Knowsley SK. Special guests for this round of training were Flameout Control and ABB, Inc.
Recognizing the importance of cutting edge testing and research, BP is providing the facility and the fuel so it can better understand and predict how LNG vapors actually disperse, in an unignited condition and how best to control the intense radiated heat when ignited. BP funded a 10,000 gallon tank truck filled with LNG to facilitate this further work and help define the most effective application rates for LNG. The tests were conducted in October just before the bi-annual BP LNG Training Workshop
Angus Fire worked out a fire test program with BP to gain the best use of this tanker load of LNG by checking whether the low application rates being proposed by some consultants and contractors from historic data, are sufficient to provide the levels of protection necessary under realistic conditions. Radiated heat flux was being measured to assess the speed with which the massive radiation levels can be reduced by application of high expansion foam from the LNG Turbex generators.
Another key part of this testing was to monitor gas levels around the spill to confirm that the flammable levels are generally close to the condensed white cloud of water vapor hanging over the containment pit.
Nearby the LNG truck waits to unload its frosty contents. Michael Moore of Flameout Control and Gerald Farnaby of ABB Automation devised a way of measuring and capturing the data to determine what is happening throughout the test. They place 15-foot vertical steel poles with three detection devices each in a straight line along the beam path of the Open Path detectors BP had installed the previous year. The straight line is on the west side of a 65 square meter concrete test plot being prepared for another man made LNG spill scenario.
The plan was to deliver a six-inch layer in the pit, somewhere between 3,000 and 4,000 gallons of LNG.
Just as important as burning LNG is the monitoring of its dissipation as the fuel is allowed to freely vent into the atmosphere.
"The test this morning is a comparison of open path gas detectors to point detectors," Moore said. "We're just going to let the gas reach a steady state and carefully monitor what happens by measuring the gas concentrations detected every second on an array of computer hardware and systems piloted by ABB." Steady state is the point at which the LNG vaporization reduces to a minimal level when all the surroundings have refrigerated down to -260 degrees F.
"We're looking at things such as egress routes and safety integrity levels to apply that information to terminals that are going to be built," Moore said. "New facilities are congregating equipment closer and closer together, making it important to understand where the gas will go."
Each of the six steel poles erected has Zellweger point detectors arranged at three different levels. The bottom row of detectors is set precisely along the same line of sight as a Zellweger open path detector, permanently installed on site.
Point detectors give readings in percent LEL (Lower Explosive or Flammability Limit). The LEL point for LNG is a 5% mixture in air, and the gas is flammable up to 15% in air. Above that it is too rich to ignite as there is just not enough oxygen to sustain fire.
By contrast, open path detectors, which measure along a line of infrared light passed from projector to receiver, measures hydrocarbon gas in LEL meters. This testing is designed to cross reference that data.
"Meter LEL just tells you there is a gas present," Moore said. "It doesn't tell you what the concentration of that gas is. Theoretically, you could have a non explosive concentration that still activates the alarm. What we're doing is setting up a battery of point detectors along the exact line of sight of the open path infrared beam stretching 15 feet."
"Combining the detectors will allow researchers to measure the vapor concentration, correlating individual LEL readings to LEL meters," Moore said. "For example, is a four meter LEL open path detector measurement equal to 15 or 80 percent LEL measured by an Infra-red point gas detector? Higher levels of point detectors in the array will not only be used to measure and verify the open path detectors, but also to attain the levels of concentration above the open path detectors to learn more about how to best place detection equipment in the future."
Backing up the open path and point gas detection equipment are "state of the art" gas imaging cameras that will allow the team to not only see how that gas might migrate and dissipate as well as conventional cameras set up and time sequenced.
"The theory is that methane heats up and rises," Moore said. "It doesn't creep along the ground like other gases. What we've discovered so far is that all the gas is remaining cold with higher concentrations around the ground than at 12 to 13 feet up where everyone is saying we should get higher concentrations."
It is believed that more data will be compiled in this event than has historically been achieved to date.
Once high expansion foam at around 500:1 expansion is applied it slows the LNG evaporation rate, but the water in the high expansion foam being used also warms the gas which forces up through the foam layer so that it goes straight up. This means lower levels of gas at ground level where ignition sources are widespread with higher levels well above the pit. Adding foam causes the gas to rise up and out of the hazard zone.
High expansion foam at the reduced application rate was generated by Angus Fire LNG Turbex generators while detectors continue to monitor the methane levels at the ground and higher levels. The foam is topped up to maintain good vapor reduction.
"Probably three times this application rate is what we think we need," said Mike Willson, Angus Fire project manager. "But we're trying to look at whether this really low rate is going to be practical in conditions where you have no ignition hazard or personnel nearby."
After the gas detection studies were completed with several foam top ups to maintain effective vapor dispersion, the detector poles were moved to prepare for ignition. LNG gas was ignited with a flaming lance to burn off the foam blanket, achieve peak radiated heat flux and then re-apply high expansion foam to assess its speed and effectiveness in reducing radiant heat down to safe levels.
A low application rate was being used based on historic test data and applied to these modern concrete pits.
Prior to the BP training facility opening last year, LNG fire testing had mostly been done in earthen or wet sand pits, which are not representative of current installations.
"Moisture in the soil tended to freeze all around the pit, so upon ignition the radiant heat melted the ice in the soil rather than heating up the pit walls, so tending to act as a giant heat sink," Willson said. "Using a concrete pit is far more realistic of actual installations as it forces the heat to build up in the pit walls and radiate outwards attacking the foam bubbles being delivered onto the LNG."
"Don't expect this fire control to be highly effective," Willson said. "We expect the intense radiated heat will carry on for much longer compared to what we've done before. It's deliberately intended to struggle and gain a sort of slow control over the fire, so we can find out where the bottom acceptability level for remote pits can be."
Not only was control of the fire about four times slower than the preferred rate, but it also caused spalding of the concrete inside the pit and on the downwind edge surrounding the pit. Clear evidence exists that the intense radiation was not being controlled fast enough to avoid damage, potentially putting personnel and plant at risk. This test was also conducted in virtually ideal weather conditions.
The big problem with such low application rates is that they have no safety allowances for rain, wind and other factors like low water pressures that will inevitably be present when an incident strikes. NFPA 11 clearly states that application rates for LNG shall be established by tests to achieve a positive and progressive reduction in radiation within the time limitations established in the analysis. "
For the next test, rather than setting the detectors in a straight line, the devices were positioned to surround the LNG pit. As the fuel vaporized more valuable data was collected
Once the detectors were removed a low flow water nozzle was used to simulate a rain storm, before lighting the LNG.
"The same low application rate was again used to try and control the fire, but the combination of radiated heat and rain prevented foam at this low rate from filling the pit and just made the situation worse, Willson said. "We had to turn on an additional LNG Turbex unit to greatly increase the application rate to bring the fire under control, which proved that a much higher application rate was necessary to take account of potential rain and wind effects."
After the testing comes the hard work of producing usable information from the research. The teams are working on this to provide comprehensive data back to the BP Group Fire Consultant Richard Coates. o