Remedial Training: Permission to Start Water Flow from Outside

Water is the natural, ­available, and effective remedy for uncontrolled fire, but we in the fire service have complicated its use through an incomplete appreciation of its actions; well-meaning, but incorrect advice; and force of habit. Recent fire dynamics research has demonstrated that we can greatly simplify our method of initial water application at structure fires by directing it into whichever building openings provide the fastest route into the involved compartment, thereby effecting immediate improvements to interior conditions. We need be no nearer to the fire than the reach of our hose streams and do not have to be inside the structure to realize these benefits.

Despite the increased ease, safety, and speed of this approach, many in the fire service are skeptical about such “shortcuts.” Notions regarding the potential for harm from exterior streams and the reportedly unique qualities of interior-placed hoselines have been drilled into generations of firefighters. Although now debunked, correcting such longstanding and systemic errors will be difficult. This article reviews the physics of water vs. fire to address lingering, yet unfounded, concerns about the “innovative” methods inspired by this new understanding.

I apologize if my arguments come across as preaching, but I have become an evangelist for the theories and methods outlined in the International Society of Fire Service Instructors (ISFSI) Principles of Modern Fire Attack. Anyone who accepts radical new knowledge that transforms their perception of the world might be similarly affected, especially when that illumination has the potential to be of immense benefit to others.

Despite my current enthusiasm, I have arrived at these convictions only after being repeatedly exposed over many years to information that contradicted my original notions. I, too, was taught that air entrained by hose streams and the immense expansion that occurs when water is converted from a liquid to steam can push fire, or at least products of combustion, into previously uninvolved areas of a building and that applying water from the exterior of a structure was of little benefit and should be resorted to only if entry was impossible.

Based on these beliefs, it was clear that water flow was best initiated from the inside, where we might position nozzles to propel hot gases toward an exhaust site—“attacking from the unburned side”—and ensure that we could reach and extinguish the seat of the fire. My beliefs were steadfast and guided my approach to understanding, performing, and teaching structural firefighting for more than four decades. They were confirmed, at least in my view, at every fire I fought or heard about, and I had no reason to doubt the theories on which our firefighting techniques were founded.

Line-of-Duty-Death Studies

It has been almost two decades since I read the first of what would be a long and ongoing series of investigative reports produced by the National Institute for Occupational Safety and Health (NIOSH) describing incidents in which firefighters were killed in the line of duty. Although many of those analyses highlighted the previously unrecognized dangers of “flow paths” [a topic outside of the scope of this article], they frequently included also the commentary that applying water earlier could have stabilized the incident and thereby prevented the fire’s progression and the firefighter fatality.

Nonetheless, we in the fire service were focused on the importance of primarily placing interior hoselines, leading many of us to dismiss this advice as impractical and ill-informed. Silly scientists, they didn’t understand our mission and resolve. But, the steady “drip, drip, drip” of these studies and their unexpected, but consistent, conclusions when “standard” interior fire attack operations “went wrong” were creating a substantial pool of evidence that contradicted the basic principles behind our methods.

Investigators with the National Institute of Standards and Technology (NIST), on behalf of NIOSH, not only performed detailed analyses of incidents, but they also often translated the collected data into computer models, graphically recreating evolving conditions and providing a dynamic means for study. Essentially, the fire behavior at an incident could be viewed as it had evolved and then be replayed again and again, allowing for detailed observations not possible in the nondigital world.

These analyses brought to light phenomena, such as the severity and speed of the spread of hot gases along a flow path and the profound effects that opening or closing a window or door could have on fire spread, which were previously unappreciated even by veteran firefighters. Despite the elegance, accuracy, and utility of these well-designed simulations, though, skepticism persisted, and the fire service at large remained unimpressed. (In our defense, no fire department has ever been called to put out a software-generated fire.)


The researchers, therefore, conducted live-fire demonstrations to test their extinguishment theories. NIST partnered with the Underwriters Laboratories–Firefighter Safety Research Institute (UL-FSRI) and the Fire Department of New York (FDNY) on a series of controlled burns on Governors Island in New York in July 2012. These extensively instrumented tests were set in nearly identical townhouses that were slated for demolition. Multiple fires and extinguishment techniques were observed and measured in an environment in which construction and contents remained consistent, providing for comparisons of multiple variables—that is, the same fires could be evaluated using different interventions and timings and then repeated in other dwellings with the same layout, construction, and furnishings to confirm or refute the findings. Such a painstakingly detailed analysis of the effects of firefighting tactics in a realistic, nonlaboratory setting had never been attempted previously. The effects of changes in ventilation, fire location, direction of water application, flow paths, and even which doors were open or shut, among other things, yielded a treasure trove of information. (For the purpose of this article, discussion is limited to findings relevant to water application. The entire experimental process and its conclusions are at

Regarding water flow, the most revolutionary outcome of these experiments, at least in my opinion, was the finding that directing a hose stream into a burning building from the exterior immediately and significantly reduced the temperatures throughout the contiguous interior areas and, even more surprisingly, did so without measurable ill effects from the flow of water itself or the production of steam that resulted. Not only were the conditions in the room in which the water was applied improved, but they were improved also in every connected space beside or above. This occurred despite an almost casual application of water, with no effort to advance the nozzles nearer to the fire.

Even more amazing to me than the unexpected widespread benefits of introducing water was that there were no detriments—no “pushing” of heat into other areas of the structure occurred, even in the areas above the fires and within flow paths. Temperatures consistently went down in the upstairs bedrooms and hallways after water was sprayed into the burning compartments below, and they did not increase from the cloud of steam. We all were sure that they would billow upward and push with it hot gases. These findings were further validated in a series of test burns in single-family structures in Spartanburg, South Carolina, two years later.

Why the contradictions of our collective expectations? It wasn’t the result of any special extinguishment agent; they used plain old water. The fire load wasn’t unrealistically small: Furnishings were typical of modern dwellings, and the burning of these synthetic materials caused a rapid progression to flashover levels. The extinguishment methods, though contrary in direction, were unrehearsed; nozzle operators were merely told where and when to flow water, not how. The results were so different from our beliefs of how water flow from the exterior would affect fire conditions that it was necessary to reassess our assumptions.

Assumptions vs. Test Realities

Why did exterior hose streams not worsen conditions? We know that hose streams entrain air, so streams directed from the exterior should increase the pressure within the structure and force smoke, heat, and steam deeper into a burning building. That did not happen at the Governors Island burns. Like many effective tactics, the success of this method lies in the details. A narrow-shaped water stream directed sharply upward with minimal nozzle movement was the unique feature of water application that allowed these exterior flows to accomplish cooling alone. This technique allows for the continued release of products of combustion (smoke and heat) and extinguishment (steam). Proving the opposite effect and recognizing that even a solid or straight stream will move about 500 cubic feet of air per minute (cfm), in those instances when the hose stream was moved back and forth across the window, blocking even briefly the exit of gases from within, smoke was pushed toward other exhaust points—in one instance, fire in a second-floor bedroom exited an open door on the first floor.

What about the increase in pressure and volume of smoke caused by the creation of steam? With water expanding 1,700-fold as it transitions from liquid to gas and the interior temperatures much higher than those required to vaporize water, this should have been a readily observed effect. The volume of a gallon of water is 0.1337 cubic feet (cu. ft.), so each gallon will produce 227 cu. ft. of steam (0.1337 cu. ft. of water/gallon × 1,700 = 227 cu. ft. steam/gallon). The minimum amount of water flowed in the Governors Island tests put about 45 gallons into a fire compartment, sufficient to create 10,228 cu. ft. of steam (227 cu. ft. of steam/gallon × 45 gallons = 10,228 cu. ft. of steam); the floor areas of the townhouses used for the burns were at most 850 square feet (sq. ft.), or about 6,800 cu. ft. (850 sq. ft. × 8 ft. ceilings = 6,800 cu. ft.). Therefore, even the least amount of water used for these tests should have resulted in almost enough steam to fill an entire townhouse, much less the one room in which the stream was directed. Despite the theoretical prediction of an overwhelming volume of steam, no increase in or movement of heat was observed after water was introduced into the burning compartments. The detrimental effects of steam did not materialize.

One reason the steam created by the water from hoselines did not behave as expected was that these were live fire tests, not laboratory experiments. Subjecting a specific amount of water to a specific temperature and then being able to accurately determine its expansion rate takes much more precision and control than can be accomplished in a burning building, no matter how much preparation is allowed. Outside of such a constrained environment, the size of that change is impossible to measure, much less predict. For instance, not all water introduced into a structure on fire becomes sufficiently heated to allow its conversion to steam; much of the steam created readily escapes through ventilation, and smoke cools and contracts as it moves away from the seat of the fire. Depending on the temperature of the gases and surfaces contacted by the water stream, a portion will be heated sufficiently to create steam, and the rest will absorb insufficient heat to progress to the gas state, remaining in liquid form on the walls or floor. Also, in the turbulent environment of a burning compartment, the same water may change to steam and cool back to liquid rapidly.

Another factor that modified the anticipated expansion of the applied water as it was converted to steam was the opposite effect that resulted from cooling the interior. Like any gas or mixture of gases, the volume of smoke is inversely related to temperature. As it absorbs more heat, it expands; as it cools, it contracts, predictably and consistently. Unlike the difficulty in measuring the steam created by a fire, calculations of the magnitude of reduction in interior gas volume are relatively easy if you know the temperatures before and after introducing water. (Temperatures must be converted to the Kelvin scale, where 273.15 equals 0°C, but then the proportional temperature change will equal the volume change.) The Governors Island experiments provided multiple examples of temperature reductions from which to compute the equal amount of gas contraction: When measured nearest the burning material, that varied from a 35 percent to a 54 percent decrease for the first-floor fires and from 56 percent to 65 percent in the basement fires. So, the introduction of water reduced the volume of smoke by at least one-third and up to two-thirds.

Why did the use of exterior streams in the Governors Island tests consistently improve conditions throughout the buildings? The profound benefits and correct methods of gas cooling were demonstrated. When the temperature in the compartment being hit by the hose stream was reduced, the heat (energy) from adjacent areas was drawn to that area, decreasing the temperatures in all connected compartments, and usually remarkably so. Note that in these tests, exterior streams did not cause the fires to go out, nor was it claimed that they should. Placing interior hoselines was still necessary to reach and extinguish the seats of the fires. Regardless, the amount and duration of improvements were substantial accomplishments. Reversing the deterioration of interior conditions, as they did, would enable us to gain additional time to search for entrapped victims and enter for direct fire attack. The decreased temperatures persisted for 90 seconds in the basement fires and for up to seven minutes in the first-floor fires. The cooling that resulted from the application of water from the exterior, even without extinguishing the fire, created predictable, reproducible, immediate, significant, and extensive improvements to interior conditions.

Exterior streams work because applying water to the interior of a burning structure cools the hot gases it encounters, causing a rebalancing of the heat energy in all communicating areas and a contraction of the volume of those gases that is greater than any expansion from the production of steam. You can prevent the propulsion of products of combustion caused by the hose stream by directing smooth bore nozzles or combination nozzles set on straight steeply and steadily into an opening. This process minimizes the associated air movement into the structure and any obstruction to the smoke and heat exhausting from the structure. The behavior of the fire in response to those water flow methods was, in retrospect, consistent with scientific principles that govern such reactions.

Applying water from the exterior was a novel approach, but the effects on the conditions within should have been predictable. Unfortunately, we had all been led to believe that other negative results would prevail, and we had never been able to accurately observe the overall reactions to our actions.

I hold no delusions that the preceding will be sufficient evidence for some firefighters, since the imperative of placing hoselines interior has been virtually “hard-wired” into our training and culture. That is the reason I am providing a more detailed explanation than “everything gets better with water,” though that is true. Taking advantage of this new research, which contradicts many of the basic ideas on which our previous tactics were based, may in some cases necessitate a reversal of our “standard” methods. If we have been doing it “wrong” for so long and have still been “succeeding,” imagine how much more effective we’ll be when our efforts are aligned with proven principles!

MARK J. COTTER is a third-generation firefighter, beginning as a volunteer in the combination department in Summit, New Jersey, in 1974. He is a volunteer engineer with the Salisbury (MD) Fire Department. Previously, he had been a member of a half-dozen other departments in roles ranging from firefighter/EMT to chief. He blogs on the Fire Engineering Training Network.

No posts to display