Browse Content by Topic:
From the Archives: Maze Meltdown: The Story of a Unique Fire
Author: Stephen L. Cutright, Fire Chief, City of Emeryville
Copyright: 9-1-1 Magazine, Feature Content
Tanker Fire Collapses Oakland Freeway Interchange
Originally Published in our August, 2007 issue.
Fires seem bigger at night.
Perhaps it’s the contrast of the bright flames against the surrounding blackness, or the billowing opaque smoke blotting out city lights, surrounding structures, and the celestial canopy. Night fires skew perspective, amplifying our sense of their depth and size. And when on occasion we are rousted out of third-cycle REM sleep, like at 3:45 in the morning, to respond to a large fire one mile from the station, the situation is ripe for cognitive and perceptual befuddlement, at least at first. Even when you know quite well the layout of the freeway interchange within which the fire is burning, even when you have an unimpeded view as you leave the fire station of 50 to 100 foot flames, the more reflective amongst us are reminded of the primal strategic admonition that fools rush in where angels fear to tread. A cautious approach was called for.
At 3:45 AM on April 29, 2007 when looking at a burning gasoline tanker truck inside the MacArthur Maze – an interchange of three major San Francisco Bay Area freeways with daily traffic of 230,000 vehicles – and the red glowing steel girders of an upper roadway directly above the fire, Captain John Arenz on Emeryville Engine 2441 chose to remain quite firmly on the side of the angels. He positioned his engine at a point where the freeway ramp split, some 75 yards away, where the 580 ramp headed southeast into Oakland and the 880 ramp headed south toward the fire. He established “880 Command” and proceeded to give his size-up report describing the fire’s location, what was involved, his strategy of letting it burn, and the resources needed to deal with the incident.
Most of the initial size-up report was factually correct. Arenz misnamed temporarily the location of the fire (saying “Eastbound 80 instead of Southbound 880) but corrected himself a few minutes later. Given the complexity of the Maze and the hour of the fire, initial confusion is to be expected. Unfortunately it took several more minutes before Oakland Fire Dispatch corrected their description of the fire’s actual location. The gasoline tanker was burning at a curve in the Southbound 880 ramp, thirty feet above ground and directly under the top “fly-over:” a double-lane bridge connecting Eastbound 80 to Eastbound 580.
Suddenly the situation changed, causing Arenz and his crew at least viscerally to question where exactly they stood relative to that strategic “angels vs. fools” boundary. With an explosion of sound, fire and violent shaking of the freeway deck upon which Engine 2441 stood, the top “flyover” ramp collapsed onto the Southbound 880 ramp, sandwiching the burning tanker truck between steel and concrete. Burning gasoline sprayed out the sides and onto the ground beneath. The bridge sections on the fire deck buckled slightly at the seams several yards in front of Engine 2441. Now the gasoline on the ground below the tanker was burning. Captain Arenz reported the new condition and started directing fire units to respond to the ground area beneath the Maze and to the adjoining freeway ramps which comprised the Maze itself. Command was passed several minutes later to Oakland Battalion Chief 2512.
Crews and media inspect the damage to the freeway structure which, due to its complex integration of three major Bay Area commuter freeways, caused severe traffic snarls in the days that followed. OFD file photo
In a manner of speaking, none of the responders had time to do the math on this fire, at least with any degree of precision. It was a typical fire in that respect, with command decisions based upon what choices were possible and effective on an incident fraught with multiple levels of risk. The decision made initially to let the fire burn itself out ended up being by far the best choice among the other response strategies available. To see this, we need to conduct a systematic elimination of what else might have been done.
Some of the alternative strategies listed below were explicitly considered while others were not. Their true importance, however, is that they comprise the balance of the cognitive field within which the strategic choice to let the fire burn was formed and subsequently re-evaluated. In the art of command, often it is as important to know the choices you’ve rejected as it is to understand the consequences of the one you haven’t.
1) Use of master streams to cool the fire: There was no water supply in the Maze freeway interchange structure. Eventually even the overhaul of this fire required the companies to connect to a distant hydrant at the filtration plant of the East Bay Municipal Utility District (EBMUD). Water use for mop-up was reasonable, given the relatively low flow volume to support hand lines. Employing master streams would have required connecting to multiple hydrants in the areas surrounding the Maze and, because of this, would have been achievable only after the fire had largely burned itself out.
The greatest problem with employing master streams to cool the fire was the environmental impact. Attacking the gasoline fire with water would have spread the fire on the ground beneath the Maze as there was no way of containing the spreading fuel and fire water run-off. This in itself would have created a large-scale hazardous materials incident encompassing contamination to soils, ground water, and the risk that contaminated fire flow run-off would reach the East Bay wetlands northwest of the Maze.
2) Quick attack: An immediate attack by the available engine companies was not feasible. In this case, the “easy math” supports the initial decision by the Incident Commander. Each engine contained five hundred gallons of water with heat absorption capacity of 8,800 BTU’s [British Thermal Units, a unit of measurement describing the heat value – energy content – of fuels] per gallon, totaling 4.4 x 106 BTU per engine, or roughly 2.2 x 107 BTU for a full fire attack of five engines. The tanker truck, on the other hand, had 8,600 gallons of gasoline at roughly 93,000 BTU’s per gallon, for a total of 8.0 x 108 BTU of heat energy on combustion. Purely on a thermodynamic basis, this meant that the fire generated over 40 times the heat energy that the first-in responders could have absorbed with all their water.
Of course, it was not so much the “easy math” that kept Captain Arenz on the side of the angels to forego mounting a quick attack. On the contrary, the scene begged for a wait and see approach. The early sight of the red glowing steel girders beneath the upper deck over the fire and its subsequent collapse convinced the Incident Commander that the basic Maze structure was at risk of further collapse. Moreover, a quick attack could not be supported by more than two other engines at best, since access to the fire was difficult for those units on the ground beneath the Maze. Indeed, the units beneath the Maze found that they could not clearly see where the fire was burning or the scope of the incident above them. When Battalion 2512 assumed command from Captain 2441, he positioned his Command Post above the fire deck, looking into the abyss, as it were, from the edge of the 580 fly-over. He, too, opted to let the gasoline burn off until the surrounding steel could be cooled by hand lines during mop-up.
3) To Foam, or Not to Foam: At some point after the fly-over collapse Battalion 2512 briefly considered using foam on the tanker truck fire, then promptly discarded the option. A foam attack on the gasoline tanker truck fire would have required obtaining foam resources from the nearby Richmond Chevron Refinery. So why was a foam attack discarded and resources never requested?
The simple answer was that by the time foam resources from Chevron would have reached the scene, the fire would have burned itself out, or nearly so. Assuredly, this would not have been the case with a tanker truck fire sitting on flat ground on a street. Common wisdom asserts that tanker truck fires need to be attacked because gasoline takes hours to burn off.
These close-up photos show the tremendous heat damage to the overpass supporting structure (note the "curtain like" distortion in the overpass I beam). A portion of the aluminum bulk fuel tank from the truck is also visible. Note that the portion of the fuel tank that was in contact with the roadway survived the inferno. OFD file photo
The rate of burning from the Maze fire was considerably greater than occurs with a tanker truck burning on a street. In the first place, this was a three-dimensional fire and the fuel could flow off the bridge and to the ground, steadily expanding the surface area of the burning fuel and consuming it faster.
Second, during the early free-burning stages of the fire while the upper deck above the truck remained intact, the heat between two freeway ramps was significant. The upper deck served to trap a portion of the fire’s heat, and while it is clear that this heat destroyed the structural integrity of the upper fly-over, we suspect that this undissipated heat also served to bring the remaining fuel within the truck’s tanks to a vigorous boil. This significantly raised the vapor production in the gasoline pool and accelerated its burn rate. Of course, no one measured the time temperature curve of this fire, but we did observe extremely active burning along the road bed from spilled fuel that, under normal “street” conditions, would have burned much more slowly.
Third and finally, when the upper fly-over collapsed upon the burning tanker truck, the tanks were flattened and heated gasoline sprayed out from under the collapse. The collapse had the effect of multiplying the surface area of the burning fuel significantly, increasing the effective burn rate.
In thirty to forty minutes the tanker truck fire was dying down, most of its gasoline having been burned away. Foam for mop-up might have helped, but it would have arrived too late to be used effectively in suppressing the main fire. More to the point, however, this was not a fire in which foam would have been effective, even if it could have arrived in time. We could not contain the gasoline within a fixed boundary or “pond” to create a non-expanding surface area and lay a foam blanket upon it to stop flammable vapor production.
The Teachable Moment
There is a certain risk that a defense of a “let it burn” strategy becomes an apologia, mere justification for an unfortunate outcome. But history is replete with examples of fires which have given us limited or no strategic options to suppress them actively. And if it is true that we exhausted our options to suppress actively the Maze fire, how can we make it go better the next time?
Certainly we can improve our dispatch communications – an expected complaint given the complexity of the Maze roadways and the plethora of frantic calls from motorists on the freeway, even at 0345 in the morning. Oakland Fire Dispatch had trouble early into the incident identifying the exact location of the fire, leading to misidentification of Maze ramps between fire units themselves. After a point, though, this was sorted out.
In the main, the communications system functioned effectively during the Maze fire. All units from the Emeryville and Oakland fire departments at the fire were linked through the Oakland Fire Department communications system, an 800 MHz trunked system managed by the Oakland Fire Dispatch Center. The dispatcher assigned a dedicated radio channel (#2) for this incident once it was determined to be a working fire in need of multiple units. In a sense, the operations on scene were fairly simple owing to the “let it burn” strategy and, although additional radio channels were available, the single dedicated channel was sufficient to handle both tactical and command communications needs.
While repairs were expected to take months, the connector re-opened in less than a month due to the efforts of a contractor with a proven track record of rebuilding damaged freeways, operating under a lucrative incentive program from Caltrans to get it done ahead of deadline. The connector collapse cost an estimated $9 million in repairs, economic impact, and Bay Bridge lost revenues. OFD file photo
The Maze fire indirectly pointed out several emergency communications weaknesses, however. First, communications between the fire units and the California Highway Patrol (CHP) were handled face-to-face in most cases, with no general radio interoperability between fire and the CHP. Conceivably an interoperable link between CHP and responding fire units might have identified the exact ramp location earlier. But the CHP, like Fire, took the conservative approach of shutting down the entire Maze and not attempting to select “safe” ramps from “unsafe” ones until the situation was clarified.
Second, the County-contracted ambulance provider, AMR, staged a number of their units in Oakland in anticipation of multiple casualties in the Maze. Word from fire units that there were no injuries (the tanker truck driver self-rescued) was delayed in getting to the AMR dispatcher or to the ambulance crews themselves. Finally, the initial 9-1-1 calls were made by freeway motorists on their cellular phones, resulting in the calls being routed to the Vallejo CHP Dispatch Center and then forwarded to Oakland. Would a direct cell routing to the Oakland PSAP have made a difference on this incident? Probably not, since the report of a fire in the MacArthur Maze got through anyway; and the Oakland dispatcher was receiving initially highly conflicting reports of the exact ramp location of the crash and fire within the Maze.
But glitches in communications aside, the most substantial way we can do better next time has little to do with emergency fire and CHP response by itself. Rather, it has to do with infrastructure. In the end, it was the catastrophic damage to this vital freeway interchange that we should most want to avoid in the future. We may assume that ill-trained drivers with bad speeding records will continue to negotiate their heavy vehicles through a dark Maze, and some will do so unsuccessfully. Given this eventuality, we must focus our efforts on mitigating our highest risk: the physical destruction of a freeway interchange. For the next Maze Meltdown, we need to build critical points in the structure to withstand and dissipate high heat, like from a tanker truck fire.
But if we are to move beyond a “let it burn” strategy on Maze tanker truck fires, we have to build engineered catchments beneath the Maze for spilled fuel and fire flow run-off; emergency water main and hydrant systems in the Maze area; and even standpipe systems with sectioned shut-off valves.
This still may be ineffective. As someone who orders men and women into harm’s way, I am biased toward caution, toward building structural defense in the structures themselves. So when it comes to rousting firefighters out of bed early in the morning, I want the structures they enter to help them do their jobs, to be on their side, firmly on the side of the angels.
Stephen L. Cutright has been the fire chief or Emeryville (CA) for more than sixteen years; prior to that he served as chief of the El Cerrito (CA) fire department for eight years. Emeryville has two fire stations serving a residential population of about 8,300 people and a daytime population of over 30,000. Emergency responses to the freeways (I-8-, I-580 and I-880) are handled within an automatic aid arrangement by Berkeley, Oakland and Emeryville.