Lessons learned from the cold storage fire at the Chicago world's fair of 1893.
Background to the World's Fair
The World's Columbian Exposition of 1893, held in Jackson Park, Chicago, was the greatest show of its type the world had ever seen. Chicago won the right to host the fair in competition with St. Louis, New York, and Washington DC, and it is said, perhaps unfairly, that Charles Dana, the disgruntled editor of the New York Sun gave Chicago the nickname "The Windy City" on account of the city representatives' continuous bragging about the merits of their bid (14). The World's Fair site stretched over 630 acres from 56 Street to 67 Street, bordered by Lake Michigan on the east side and by Stoney Island Avenue on the west, with additional attractions on the Midway Plaisance, a strip of parkland between 59 Street and 60 Street to the west. The exhibition ran from May 1 to October 30, and was attended in total by over 27 million visitors (just over half the population of the United States at that time). The World's Fair claimed many notable "firsts," including the introduction of hamburgers, Aunt Jemima[R] pancake mix, Hershey's[R] bars, and Juicy Fruit[R] chewing gum to the American public.
The Exposition's organizers wanted to provide a showcase for Chicago's achievements in science, engineering, and the arts to the rest of America, and, in turn, demonstrate the emergence of America as an international power to the rest of the world. Parisians had unveiled their famous tower, designed by Gustave Eiffel, in 1889, and Chicago required something even more dramatic. A local mechanical engineer, George Ferris, designed an amusement ride in the form of a giant wheel, was 80 m (264 ft) in diameter, that could carry 2160 passengers. Although it was only one-quarter the height of the Eiffel Tower, and was dismantled at the end of the fair, the Ferris wheel became the internationally recognized symbol of the fair.
To maintain an air of respectability in the main exhibition areas, all of the amusements and side shows were located on the Midway Plaisance. This included the Ferris wheel and, next to it, the only ice-related exhibit in the entire fair--the De La Vergne Ice Railway, which is shown in Figure 1. Several other amusements that had not been granted concessions by the organizers, including Buffalo Bill's Wild West Show, were located to the west of Stoney Island Avenue south of the Plaisance.
[FIGURE 1 OMITTED]
Despite the importance of refrigeration to the key Chicago industries of stockyards, railroads, and brewing, there were no exhibits highlighting this new technology in the main exhibition area. This oversight was severely criticized by the editor of the journal, Ice and Refrigeration, who wrote in the December 1892 issue that:
... the exploitation of the science has been so hedged about by impracticable conditions for the exhibition of its appliances that an intelligible exposition of the subject, in which the whole world, in all countries and in all climes, has so deep, so profound and so direct an interest, is only possible under the conditions of a 'concession', bought and paid for in cash by exhibitors who stand in the same grouping and class with the peanut vendor and the pop-corn stand (2).
The nearest thing to a refrigeration display was the cold storage building, erected by the Hercules Iron Works of Aurora, Illinois, to provide warehousing and ice-making facilities to the fair. As Ice and Refrigeration's editor observed:
... when the problem of feeding the multitude of spectators--expected to average 175,000 daily--came to be considered, then the management were compelled to call on this science to lend its aid to this immense undertaking ... The refrigerating and ice making department of the Exposition therefore, while it was in the nature of an after thought, came of necessity to be very properly regarded by the directors as second only in importance to the provisioning department - an absolute essential to that department in fact ...
The cold storage building was funded by the directors of the Hercules Iron Works and through additional subscription, and it was located on the east side of Stoney Island Avenue just south of the 64th Street entrance, at the back of the Central Railroad Station. A map of the fairground, highlighting the location of the cold storage building, is shown in Figure 2, and the same location on a modern map is given in Figure 3.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Background to the Study
This reappraisal of the cold storage fire at the World's Columbian Exposition was prompted by an article published in 2005 by Professor Jonathan Rees, Associate Professor of History at the University of Southern Colorado. Professor Rees suggested that the use of ammonia in ice plants of the late 19th century was fundamentally unsafe, and that this was well known by the builders and operators of these refrigeration systems. He cited the cold storage fire at the World's Columbian Exposition as typical of installations that "exploded," causing injury and death, and postulated that the industrialists of the time chose to ignore the hazards of ammonia use in order to sell unsafe systems, writing:
Through their campaign of misinformation, partisans of ammonia distorted the free market in refrigeration equipment to favor a dangerous, inefficient technology over other, safer alternatives (3).
He also quoted the Chicago Inter-Ocean, which had reported that there had been a "deafening explosion of the ammonia pipes" (4). The difficulty with Professor Rees's analysis of the situation was that ammonia does not behave in the manner he described. These inconsistencies warranted a more detailed technical appraisal of the disaster, with a view to gaining a better understanding of what actually happened.
DESCRIPTION OF COLD STORAGE FACILITY
The plans for the World's Fair had not included a cold storage facility within the main exhibition area, but it was recognized that the many food concessions, restaurants and bars would have a need for ice for food storage and beverages and a supply of frozen and fresh produce on a daily basis. The Hercules Iron Works was selected by the World's Fair organizers as the preferred supplier of the facility. The Hercules factory was in Aurora, Illinois, about 40 miles west of Chicago, and they also had sales offices in Chicago and New York. In the preceding four years Hercules had built a strong reputation as supplier of efficient, reliable ice-making plants, including installations in Tennessee and Louisiana described in most favorable terms in testimonial letters published in Ice and Refrigeration March 1893 (5). The Hercules "Model Ice Plant" used a steam engine powered by oil- or coal-fired boilers to drive twin-cylinder ammonia compressors. This is shown in Figure 4. The ammonia was used to chill calcium chloride brine that then circulated to a block ice tank. The plant installed for the World's Fair is described in Ice and Refrigeration (2) and in an article in Scientific American that was published two weeks after the fire (6).
[FIGURE 4 OMITTED]
The building was five stories high, designed by architect Franklin P. Burnham in the style of a Moorish palace, as shown in Figure 5, with a statue of Christopher Columbus holding a globe at the main entrance. The main building was 255 by 130 ft with plain white walls broken only by windows on the top floor, which was equipped as a skating rink. The towers on each corner of the building rose to 115 ft, with the central tower, which housed the boiler flue, rising a total of 191 ft above street level. The store roof was about 60 ft above street level, and the gallery around the central tower was at a height of about 120 ft. Access to the gallery was provided by a staircase from the top floor, but there was no means of ascent from there to the top of the chimney stack.
[FIGURE 5 OMITTED]
In the center of the building, covering two stories, was the main machinery room, measuring 80 by 80 ft and containing three ammonia compressors, steam engines, air compressors, and dynamos to provide electrical power to the building. Behind the machinery room was the boilerhouse, containing multiple boilers with a total capacity of 700 hp and tanks for collecting distilled water from the boilers for the production of clear block ice. On the south end of the building, apparatus for making plate and block ice was installed and was in full public view. Behind the ice-making room was an ice storage chamber measuring 100 by 40 ft, and capable of holding 1000 tons of block ice. The north end of the building was divided into various sizes of chill and frozen storage chambers over four floors, giving a total of 600,000 [ft.sup.3] of storage. A floor plan of the building, taken from the December 1892 edition of Ice and Refrigeration is shown in Figure 6. The third and fourth floors in the center and southern portions of the building, above the boiler room, engine room, and ice plant, were fitted out as offices and apartments for Hercules staff and their families. On the top floor a skating rink had been created, measuring 208 by 54 ft (approximately 64 by 16m). The inclusion of an ice rink in the design was said to be a late addition to the plans, perhaps prompted by the lack of other refrigeration exhibits in the fair, with the notable exception of the De La Vergne ice railway on the Midway Plaisance. The rink was located on the east side of the building, with a broad promenade around it. The ice rink ran for most of the length of the building, leaving about 25 ft at each end, but it only covered about half the width of the building. The rest of the floor area was intended to be used as a cafe and smoking room.
[FIGURE 6 OMITTED]
The whole building was intended to provide a showcase for new technologies and, consequently, was rather more complex than necessary for the storage and distribution of provisions to the fair. Three types of ice makers were installed--the plate system, the can system using condensed steam, and the can system using de-aerated water. The cold storage rooms, which covered four floors at the north end of the building, also used three methods of cooling. Air was circulated from ammonia-cooled evaporators on the fifth floor, feeding some of the chambers on each floor, while other rooms were cooled either by direct expansion ammonia coils or by recirculated brine.
The building insulation comprised double air-gaps with wooden boards and 2 X 6 in. studding. The boards were lined on both sides of the partition with "insulating paper." This would provide no insulating effect, but acted as a vapor barrier to prevent moisture ingress to the cold space. On the outer side, slabs of "staff"--a mixture of plaster of paris and hemp fiber--were nailed to the studs, providing a smooth white outer finish. The arrangement of this insulation is shown in Figure 7, taken from the De La Vergne catalog (1).
[FIGURE 7 OMITTED]
The construction of the boiler flue had already attracted a great deal of criticism and had caused a few minor fires in June, resulting in the cancellation of most of the insurance policies on the building. The main complaint was that the cupola around the top of the iron flue was intended to be made of wrought iron and lined with asbestos to protect the wooden tower from sparks and heat, but this protection was missing (7). Despite several warnings and instructions to rectify this defect, the remedial work was never done. A further complaint was that the boilers were generally undersized for the duty required of them (7).
EVENTS OF JULY 10, 1893
A detailed eyewitness account of the disaster is provided by the unnamed correspondent for Scientific American who happened to be in the building researching a report on the novelty of the ice rink when the fire broke out (6). Although the cold store had been in service since April and the ice plant had been operational since the beginning of the fair in May, the ice rink was just in the process of being commissioned and the brine had only been charged to the pipes in the floor on July 8. The first flood of the ice rink floor took place the morning of July 10, and by early afternoon half the floor surface had been frozen satisfactorily.
The alarm sounded at 1:30 p.m. when a small fire at the top of the flue stack was spotted. The responders came from the World's Fair fire service and the local fire station, which was only a few blocks away. As on previous occasions, about 12 firemen climbed onto the roof of the cold store and climbed the staircase from the ice rink on the fifth floor to the gallery of the central tower. From there they nailed boards to the walls to create footholds and climbed up to the narrow ledge that formed the roof of the gallery. However, shortly after reaching the ledge above the gallery, just below the summit of the tower, there was a sudden outburst of flame below them, preceded by a small puff of white smoke (7). The fire at roof level spread very rapidly and cut off the escape route for the fireman up the tower. A few were able to slide down the ropes used to drag the hoses up, but the ropes and hoses were quickly burned through, and the rest of the group were left with no means of escape. One by one they jumped the 60 ft from the gallery to the main roof, but all were killed or seriously injured by the fall.
INTERPRETATION OF THE EYEWITNESS DESCRIPTIONS
Perhaps the most compelling of all the journalists' reports of the fire is from the unnamed writer for Scientific American, who was in the building at the time the fire alarm sounded, and who witnessed the development of the tragedy first hand. His report is notable because it does not use the term "explosion" at any time, and does not mention the smell of ammonia. He says that the team of firemen who responded to the alarm
were engaged with ropes in drawing up the hose pipes, to extinguish the flames above them, when, all of a sudden, to the horror of the spectators, the flames burst fiercely out from the lower part of the tower, far below where the firemen were at work (6).
The Chicago Daily Tribune provides a more detailed report, with many details consistent with an eyewitness account (7). Their version of the outburst of flame describes how the fire started in the cupola at the top of the stack, but burned down to the wooden columns below the dome. Pieces of burning timber were seen falling from the top of the tower- some outside the cladding and some between the stack and its pine cladding. The puff of white smoke "that came from near their feet" was immediately followed by tongues of flame bursting out from the tower, below the platform occupied by the firefighters, but well above the roof of the main building.
The reports from further afield are rather less believable. The New York Times reported that "Chief Sweeney" [Swenie] of the Chicago Fire Department attributed the blaze to "an explosion of ammonia used in the manufacture of ice" (8). The same paper also stated that the central tower that housed the smokestack was also used as "ventilator and smokestack, the ammonia fumes from the ice machines escaping through it." Likewise, the Los Angeles Times provides a very graphic but rather overstated account, lacking eyewitness details. The Los Angeles Times article states that up to 75 firemen were on the roof of the cold storage building, that 30 firemen had been up the tower when the flames broke out, and that all of them had been killed. In addition, three women and several clerks from the cold store's offices were said to have been crushed by falling masonry. It also reports that "a poor fellow, Murphy by name, was taken down but died on the way to the hospital." Although Bernard Murphy, a boilermaker, was one of the victims, he was inside the building and was killed on the spot when the tower collapsed (8). It seems that the Los Angeles Times reference is to Captain James Fitzpatrick, who led the team of firefighters up the tower but was badly injured when he leapt 40 ft from the end of a half-burned rope to the roof shortly before the tower collapsed. He was rescued from the roof by Fire Marshal Edward Murphy, acting Chief of the World's Fair fire department, and two others, named by Marshal Murphy, as Captain Richard Kennedy and Hans Rehfeldt. Fitzpatrick died at the emergency hospital.
Ice and Refrigeration carried a full report of the fire in the August 1893 issue, quoting extensively from an unnamed "contemporary" (10). Unfortunately his details also do not fit the eyewitness accounts. The Ice and Refrigeration report states that Captain Fitzpatrick was one of the first to jump from the tower, when in fact he was the second last, having delayed his leap to ensure that all his team had evacuated. According to the reports of the Chicago Daily Tribune (20), the last man was a young electrician, Norman Hartman, who had been wiring up large electrical signs on the tower when the fire broke out, and climbed with the firemen to the gallery level, then found his escape route cut off by the fire. By the time Fitzpatrick and Hartman tried to escape, the last rope was only 20 ft long. Fitzpatrick slid down the rope and launched himself northward, managing to clear the flames but breaking his right leg in the process. However, as Hartman reached the end of the rope, the tower collapsed over him before he could make the leap. The rather more fanciful New York Times report goes on to say:
The men had made the terrible leap as the only possible chance of saving their lives, and landed on the tar and gravelled roof, only to be so solidly imbedded in the sticky, yielding composition that they could not have extricated themselves had they been granted the strength to do so. Thus imprisoned in this quicksand on the roof, they were held firmly until the huge tower crashed upon them and sent them down into the fiery furnace underneath (8).
There is no suggestion in any of the other reports that any depth of molten tar existed on the roof, even at the height of the fire, and the able-bodied firefighters had no trouble crossing it. The clear reason for the lack of mobility of those who jumped between 40 and 60 ft from the tower was the injuries they sustained in the fall.
There are various comments in the press immediately following the tragedy denouncing the building as a fire trap. Marshal Murphy, the rescuer of Captain Fitzpatrick, is said to have stated a week before the fire that he did not think the building would last until the end of July.
EVIDENCE FROM THE INQUEST HELD IN CHICAGO ON JULY 13-18, 1893
An inquest was held, under Coroner McHale, for the 16 known victims, including one unidentified corpse for the purposes of determining cause of death and, where possible, apportioning blame for the fatalities. As witnesses were called from the staff of Hercules Iron Works, the fire department, and the management of the World's Fair, several contradictory strands of detail emerged. Walter Donaldson, manager of the cold store, testified that the fire was already burning in the lower portion of the tower before the firemen arrived. He asked them to cut into the staff at the base of the tower and pour water on the flames below, but his advice was ignored and the firemen climbed to the top of the tower. John Skinner, President of Hercules Iron Works, testified that one of his staff, A.J Brand, had also tried to persuade Captain Fitzpatrick to cut into the base of the tower, but he was punched in the face and knocked down. Skinner also testified that 38 minutes elapsed from the time the alarm sounded to the time the water was turned on for the hoses, and suggested that this delay in fighting the fire at its source was the reason that it was able to take hold so completely (17). Walter Donaldson stated that Murphy and Fitzpatrick arrived together and started sending their men up the tower with ropes to haul up hoses. Murphy gave instructions to the team on the ground not to turn on the water until he called for it. J.D. Alsup, the Construction Supervisor for Hercules, also testified that the firemen ignored advice about fighting the fire below. James Anderson, chief carpenter, said that he had pulled off one of the boards around the shaft on the fifth floor, under instruction from Mr. Alsup, and found flames coming up from below. He called for water and shouted to the firemen up the tower to evacuate, but was ignored on both counts. M.W. Smith, Chief Engineer for Hercules, corroborated the accounts of Donaldson, Alsup, and Anderson, and said that he thought the fire had started between the third and fourth floors (18). This suggestion is supported by two women, Mrs. Harris and her sister-in-law, who had been in one of the apartments on the fourth floor. Without hearing any alarm or call for evacuation, they felt that it suddenly got very hot in the apartment and decided to go outside. They noted that the floor had been particularly warm-almost too hot to touch. The fire engines arrived just as they were leaving the building (17).
In contrast to the evidence of the cold store employees and the two women, the members of the fire department who were questioned at the inquest maintained that they were only aware of the fire at the very top of the stack. Marshal Murphy stated that:
The fire was burning on the extreme top of the cupola or chimney and did not extend over six inches over the top of the chimney. That was the only fire to be seen and there was no smoke from the fire whatever (18).
Murphy's account suggests that Fitzpatrick was already up on the ledge above the tower gallery when he arrived on the roof, and was calling for a ladder to get closer to the top. Once the ladder was hoisted and positioned, the water was turned on. As soon as the water hit the flames smoke appeared, followed by an "explosion," and flames burst from all sides of the tower below the gallery.
A remarkable discrepancy appeared in the Fire Department's evidence during the inquest, related to the rescue of Captain Fitzpatrick. Marshal Murphy's account states that he was on the roof when Fitzpatrick jumped. Fitzpatrick broke partway through the roof and, although lying clear of the flames at the base of the tower, he was trapped and being scorched by flames coming up from below. As Murphy tried to free him, the tower collapsed and covered them both in debris. Murphy struggled clear and called for assistance from two firemen, Williams and McIntire, as well as a Columbian guard, who were all on the roof. Together they moved the debris and managed to pull Fitzpatrick free. As they got to their feet, there was a large explosion that lifted the roof at the center and knocked Murphy to his knees. The roof started to subside, so Murphy left Fitzpatrick and slid down a hose on the north side of the building. He went to the east side and sought help from two firemen whom he identified as Rehfeldt and Kennedy. They went up a turntable ladder, found Fitzpatrick, attached a rope to him, and lowered him to the ground. He was taken immediately to the emergency hospital, but died there shortly after arrival. This account differs significantly from that given by Fire Chief Swenie, under oath at the inquest. Swenie stated that the two firemen who rescued Fitzpatrick from the roof were Lieutenants Miller and Barker of Fire Companies 52 and 16, respectively, acting under the instructions of Marshal Murphy. Miller and Barker both seem to have confirmed this account, which was also witnessed by Officer Charlton of the Chicago Police Department (19). Charlton reported that he later saw Fitzpatrick's body at the hospital, and recognized it because the right side of the face was badly burned and half the moustache was missing. Murphy stoutly maintained that he could not have been assisted by Miller and Barker, as Companies 16 and 52 were not called to the blaze until after Fitzpatrick had been sent to the hospital; however, both men later testified that they had been off duty and visiting the fair when the alarm was raised, and so joined the effort voluntarily before their colleagues arrived.
SIFTING THE EVIDENCE
What Can Be Established Beyond Doubt?
The fire broke out at the top of the boiler flue, where the original design had not been followed. The wooden tower facing was not protected from the hot fumes and possible ash fragments coming out the chimney. This construction defect had not been rectified, despite two previous fires in the same place. However, the emergency responders clearly believed that the fire could be dealt with quickly and without danger. During their efforts to establish a vantage point to fight the fire, flames broke out below them, blocking their escape route. The high loss of life was a direct result of their rapid ascent of the tower, using boards nailed to the tower to create makeshift rungs rather than waiting for ladders to be positioned.
The destruction of the cold store seemed to progress in four stages. The initial small fire, reported just after 1:30 p.m., was followed by a second outbreak, still within the tower but at a level below the gallery. This probably occurred at about 2:00 p.m. At this point the flat roof of the store was still intact, and firefighters on the roof were able to play hoses onto the lower portion of the tower. A photograph taken at this stage is shown in Figure 8.
[FIGURE 8 OMITTED]
The firemen on the ledge above the gallery initially hesitated, waiting for ladders to be brought up. As they waited, the flames in the tower below them became more intense and the possibility of escape by ladder was ruled out. They were ushered to the north side of the tower where the heat was less intense. By then it was not possible to slide down hoses or ropes because the fire in the tower below them was too strong. One by one they slid as far as they could and then jumped. By this time the roof area around the tower was also burning, and several men, failing to clear the damaged portion, fell through the roof. The last two on the ledge were Fitzpatrick, who had ensured all his men jumped first, and Hartman, the young electrician who should not have been there at all. Hartman was unwilling to jump and so hesitated. Fitzpatrick tried to persuade him down, but eventually had to leave and tried to make his escape by sliding down a rope and then pushing off to the north. Hartman was trying to descend the rope when the tower collapsed, partially covering Fitzpatrick in debris. When the tower collapsed the flames spread rapidly throughout the building. From this point onward there was no question of quenching the flames. The fire department retreated and left the building to burn, concentrating their efforts on ensuring that adjacent buildings were protected and that the fire did not spread to the neighboring hotels or the rest of the fairground. By 4:30 p.m. the building was destroyed and the search for bodies had commenced.
The evidence of the cold storage employees and the women visitors suggests, however, that the fire had already taken hold on a lower floor, perhaps even before the alarm was raised. If this were the case, then it is clear that there were no obvious signs of a lower fire when the emergency responders arrived.
What Can Reasonably Be Deduced?
As reported in Scientific American, the ice rink had only been put in service on the day of the fire. It therefore seems likely, as ice was being made for the first time, that the heat load on the refrigeration system and the steam demand were at higher levels than had previously been achieved. Press comments about the inadequacy of the building report that the boilers were rather undersized. Certainly the boiler output and refrigeration capacity reported in the Ice & Refrigeration review in December 1892 were somewhat scaled down: the installed compressor capacity was 240 TR rather than 300 TR, and the boiler output was 700 hp rather than 800 hp.
It is also clear comparing the December 1892 description with later accounts that the fifth floor plan was substantially modified to accommodate the addition of the ice rink. This modification may have included the removal of the cooling coils associated with the air circulation system, and a reduction in scope, or possibly elimination of the air circulation method. However, even without the air ducts required to feed cold air to the storage rooms, there were plenty of air gaps within the insulation construction of the walls and floors. It is likely that the collapse of the tower is what spread the fire horizontally within the building, forcing burning embers through the air gaps in the insulation and causing rapid destruction of the fabric of the whole building. This is probably what was described as an "explosion" in the press. It should be noted that Murphy, Kennedy, and Rehfeldt (or Miller and Barker) must have rescued Fitzpatrick from the roof after the collapse of the tower, by which time the whole building below them was on fire.
In fact, there appears to have been three "explosions" in the progression of the fire. The first immediately followed the first jet of water applied to the top of the tower and produced a "puff of white smoke" below the gallery, immediately followed by the eruption of flames all round the lower portion of the tower. The second explosion caused a rapid spread of fire throughout the upper floors of the building, probably caused by the rush of hot air and embers following the collapse of the tower. The third explosion, reported by Murphy, occurred while he was trying to move Fitzpatrick to the edge of the roof. It caused the roof to lift and settle, but did not result in any immediate structural collapse in a building that was already severely weakened. It is likely that this was deflagration of ammonia on the lower floors of the building, possibly on the condenser flat that was immediately above the boiler room on the second floor. If so, then it occurred after the initial fire, after the spread of flames within the tower, after the collapse of the tower, and after the further spread of flames throughout the upper part of the building.
What Other Alternatives May Exist?
The firemen would not have been so quick to clamber to the ledge up the outside of the tower with their makeshift rungs if, as suggested by Ice & Refrigeration (10), it was already obviously on fire, or if there was evidence of burning fragments falling down inside it. It is possible that the column was protected initially by a cover or platform at the top and the fragments could only fall down once the cover had been burned away, which happened while the firemen were on the gallery. As this had not happened in the previous fires, there was no perception of the danger of being cut off by the fire.
It may be that there were in fact two fires: one at the top of the stack and one much lower down, perhaps on the second or third floors. If so, then it is feasible that they were both caused by the high load on the boilers and ignited at about the same time. The report in Scientific American (6) praised the boilerman for drawing the fires of the boilers before evacuating the building and, thus, avoiding a much more damaging boiler explosion. If the lower outbreak had been in the boiler room, it would have been obvious, but at this stage there seems to have been no visible fire at that level, and no substantial ammonia leak. When the fire department responded to the alarm, which may have been raised in response to the lower fire, they reacted to what they saw from the outside of the building-the by-then familiar sight of a small fire at the top of the stack.
What Can Be Discounted?
It would not be possible for the fire at the top of the stack to cause an ammonia leak at the bottom, on the other side of a fire proof wall. It is not possible for the ammonia inside the refrigerating system to "explode" by combustion, as suggested by the fire chief, reported in the New York Times and the Chicago Inter-Ocean, and noted by Professor Rees. The upper flammable limit for ammonia is 28% v/v, which means that there would need to be sufficient air in the pipes to account for 72% of the volume. As the refrigeration system would have failed to function with such high levels of noncondensibles in the condensers, this explanation can be discounted. "Explosions" can also occur when pipework and components are pressurized beyond yield and they burst without any combustion. The destructive power of these bursts is illustrated in Figure 9, which shows the remains of a service truck after a 60 kg (132 lb) ammonia cylinder burst due to hydrostatic pressure.
[FIGURE 9 OMITTED]
In refrigeration systems, this can occur for two reasons. In a closed volume with gas and liquid, a temperature rise will cause a rise in the saturation pressure of the refrigerant. In a fire situation, the temperature could rise to levels above the design pressure of the components. This is the main reason that pressure-relief valves are now required by safety codes, but they were not introduced to pressure systems in the United States until 1915 after a series of boiler explosions, and so it is certain that they would not have been fitted to the Hercules plant. If liquid is trapped in a pipe or vessel with no gas space and is then heated, then the thermal expansion of the liquid will generate very high pressures for even a relatively small temperature rise. This can cause excessive pressure at ambient temperatures, but requires that the pipe or vessel has been filled with liquid and then isolated from the rest of the system. This can happen in modern plants with non-return valves and solenoid valves, and is the reason that pressure-relief valves are fitted to liquid lines on pumped systems. At the Hercules plant it is unlikely that there were any automatic solenoid valves or non-return valves in the liquid pipes, so hydrostatic pressure is not likely to have occurred during the Chicago fire.
The design of the engine room included a public viewing gallery to show off the new machines. It seems highly unlikely that there would have been a significant smell of ammonia in the engine room during normal operation. Ammonia causes a smell at 0.0005% in air (5 ppm by volume), and is unbearable at 0.05% (500 ppm), but it is not flammable at concentrations below 15% (150,000 ppm)-three hundred times higher than the unbearable concentration. It is also very unlikely that the chimney tower was used to ventilate ammonia fumes from the engine room as suggested by the New York Times, as the bottom of the boiler flue was in the boiler room on the other side of a fire-proof wall from the ammonia compressors.
From the review of the technical journal and general press reports, and from a study of photographs of the site before and after the fire, the following factors have been identified as contributing to the disaster.
* Chimney Design. The top of the chimney was not installed in accordance with the approved drawings, according to several contemporary reports. The iron chimney stack stopped a few feet short of the top of the decorative wooden facade.
* Building Construction. The face of the building looked like marble, but was in fact staff slabs, nailed to wooden frames. Insulation was by air gaps lined with wooden boards faced with "insulating paper," a form of vapor barrier. All of this material was highly combustible, and would have been dried out after two months of cold store operation.
* Air Circulation. It is not clear whether the air circulation system described in Ice & Refrigeration was installed. Cost-saving measures may have caused it to be dropped, or the additional space required for the ice rink, cafe, and smoking room on the fifth floor might have precluded its installation. If it were there, even in a scaled down form, it would have made the spread of fire throughout the building much quicker and more difficult to prevent.
* Boiler Size. If the boilers were reduced in output from the original design, but an ice-rink was then added to the scheme, then it is likely that they were rather small for the duty, and they may have been stoked excessively to generate sufficient power to cope with the high heat loads on the plant.
* Equipment Style. It is easy to look at the design of equipment and conclude that the fire occurred because the cold store was cooled by a coal-fired plant. This was normal for the time, and there were many cold store and ice plant fires (3), but we should not be too complacent. The majority of cold store fires today are caused by electrical faults or "hot works"-welding or brazing during installation or repair.
* Inadequate Emergency Resources. There was no means of access to the top of the tower despite the previous fires, and no ladders stored onsite for the purpose. The water pumps onsite could not generate enough pressure to hose the top of the tower, even from roof level. The fire department was not equipped with nets to catch people jumping from buildings.
* Late Commissioning. Although the ice plant had been in use since April, and the cold store complex since May, the late addition of the ice rink was only being completed on the day of the fire. This meant that the plant was working harder than ever before, with equipment used in ways that had not previously been experienced. It is important that commissioning is not just viewed as "setting to work," but is used to confirm that the installation performs as the designer intended. This was clearly not done in this case.
* Over confidence. In the case of the cold storage fire, there was overconfidence in the operators who ignored previous warnings about the fire hazard. There was also overconfidence in the fire department, who did not wait for adequate equipment to arrive and did not keep a close fire watch on the foot of the tower. Daniel Burn-ham, Head of the World's Fair Construction Department; Marshal Edward Murphy, Fire Chief; and two of the senior officers of the Hercules Iron Works, John Skinner and Charles McDonald, were held over by the Coroner's jury at the inquest to appear before the Grand Jury. However, the Chicago Daily Tribune reported on August 6 (12) that all four had been absolved of all blame. The New York Times on (13) summed up on July 12 the feeling of over-optimism that had preceded the disaster:
At the northeast corner of the building was a mountain of smoking timbers, from the middle of which protruded another sign, announcing with ghastly sarcasm: 'This building strictly fireproof.
LESSONS FOR TODAY
* The Effect of Cost Cutting. There is no doubt that penny-pinching played a part in the cold storage disaster. The originally approved plans were not followed, presumably to reduce the cost. Emergency equipment was not available (for example ladders), and what was there was inadequate for the task (for example water pumps). These lessons have not been learned: contractors and end-users continue to taken uncalculated risks by skimping on safety provisions in their projects, or by ignoring seemingly unimportant design details in order to achieve a cheaper result.
* The Difficulty of Accident Investigation. There are two problems apparent in any accident investigation: sometimes there is not enough information, but at other times there can be too much. Each witness statement must be carefully weighed. The investigator must determine whether the witness is credible, whether his statement makes sense, and whether it matches the stories told by others. Often it is possible for two apparently contradictory statements to be reconciled once the whole story is understood. It is also possible that both contradictory witnesses are wrong and the truth lies elsewhere. Eyewitness accounts are particularly important. Not only do they tell you what they saw; they will also fill in what they heard, smelled, and even felt. Equally important they will tell you what they did not experience; for example, there were no reports of the smell of ammonia in the eyewitness accounts. Perhaps the most important part of an eyewitness account is that it may include significant observations without any understanding of their meaning. For example, the observation of a puff of white smoke from the bottom of the tower immediately before the flames burst forth suggests that the point of ignition may have been within the tower, not further away; the smoke from burning staff and timber was thick and black, as shown in Figure 8. That the smoke could be seen by witnesses standing on the road indicates that it was partway up the tower, not elsewhere on the roof. Alternatively it may have been steam, caused by rapid vaporization of the first jet of water to be played onto the tower. This could have caused an air pressure wave that would have spread the flames in the lower portion of the tower, causing them to emerge from the stack all round the base of the tower.
* The Dangers of Ignoring Warnings. There were plenty of opportunities for the staff of the Hercules Iron Works to resolve the problems with the chimney stack before July 10. Several insurance policies had been cancelled in the weeks before the fire as a result of the previous fires in the chimney stack (7). Whatever caused the small fire on July 10 to spread so terribly, it is clear that modifications to the top of the stack after the previous blazes would have prevented the start of the chain of events or would have prompted the emergency responders to attack the real seat of the blaze, further down the stack. It is important to instill a safety culture that treats near misses as seriously as casualties. Near misses need not be at the facility in question-any accident from any sector of industry can serve as a timely warning if analyzed correctly.
The tragedy at the cold storage building of the Chicago World's Fair on July 10, 1893, was caused by a combination of human error in designing the building, in executing its construction and commissioning, and in responding to the initial fire alarm. It is clear from a detailed analysis of the spread of the fire that the use of ammonia in the refrigeration system was not a contributory factor and that the presence of ammonia onsite did not hamper the firefighting or rescue efforts, nor did it significantly accelerate the destruction of the building once the firefighting had been abandoned. Twenty-first century refrigeration systems are built in compliance with much more stringent safety codes, but the hazards of fire, safe access to high locations, and excess hydrostatic pressure still claim lives in the refrigeration industry and the emergency services.
The author would like to acknowledge the assistance of Star Refrigeration, Ltd., of Glasgow, UK, in supporting this study, and to recognize the initial motivation and subsequent encouragement and support given by Professor J. Rees. Thanks are also due to D. Dettmers and B. Nagengast for further encouragement and assistance in finding much of the source material.
It is also appropriate to acknowledge the victims of the tragedy: Captain James Fitzpatrick, Captain James A. Garvey, Captain Burton E. Page, Lieutenant Charles W. Purves, Mr. William Denning, Mr. Louis J. Frank, Mr. John H. Freeman, Mr. John A. Smith, Mr. Philip J. Breen, Mr. John McBride, Mr. Paul Schroeder, Mr. John Cahill, Mr. Ralph A. Drummond, Mr. Bernard Murphy, and Mr. Norman M. Hartman.
They are commemorated by a monument at Oak Woods Cemetery, Chicago, shown in Figure 10. One corpse included in the inquest and the official death toll of sixteen was never identified.
[FIGURE 10 OMITTED]
Their colleagues, injured in the fire but not named on the memorial, and members of the public also injured are also remembered: Captain Thomas Barry, Frank Bielenburg, H. Breckenridge, John Davis, T. Donahon, Frank Faulkner, W. Fisher, G. French, Fred Goetz, G. Haman, Captain Richard Kennedy, Captain Kenyon, William Lenahan, Dennis Lynch, William Lynch, Gus Martin, William Mahoney, John McIntyre, Mrs. D. Moore, Marshal Edward Murphy, M. Murray, Sigmund Nordstrum.
(1) De La Vergne Co. 1897. Mechanical Refrigeration and Ice Making, New York
(2) Ice & Refrigeration, December 1892. 3(6):427-432
(3) Rees, J. 2005. "I Did Not Know ... Any Danger Was Attached": Safety Consciousness in the Early American Ice and Refrigeration Industries Technology and Culture - Volume 46, Number 3, July 2005, pp. 541-560
(4) Chicago Inter Ocean, 11 July 1893:1
(5) Ice & Refrigeration, March 1893. 4(3):177
(6) Scientific American, 22 July 1893. 69(4):52
(7) Chicago Daily Tribune, 11 July 1893:1
(8) New York Times, 11 July 1893:1
(9) Los Angeles Times, 11 July 1893:1
(10) Ice & Refrigeration, August 1893. 5(2):89-91
(11) Dolomiten Tagblatt der Sudtiroler, 23 January 2008:1
(12) Chicago Inter Ocean, 6 August 1893:8
(13) New York Times, 12 July 1893:5
(14) Chicago Tribune - Lifestyle section, 7 December 2004:1
(15) Map adapted from Paul V. Galvin Library Digital History Collection Map of the Buildings and Grounds of Worlds Columbian Exposition. http://colum-bus.gl.iit.edu/
(16) Map adapted from Google maps. http://maps.google.co.uk
(17) Chicago Record, 17 July 1893:1
(18) Chicago Daily Tribune, 18 July 1893:3
(19) Chicago Daily Tribune, 17 July 1893:1
(20) Chicago Daily Tribune, 12 July 1893:1
Bancroft, H.H. 1894. The Book of the Fair: An Historical and Descriptive Presentation of the World's Science, Art and Industry, As Viewed through the Columbian Exposition at Chicago in 1893. New York: Bounty.
Further details about the World's Fair of 1893 can be found at the following Web sites:
The Chicago Historical Society. http://www.chicagohistory.org. The Illinois Institute of Technology. http://columbus.gl.iit.edu/.
Andy Pearson, PhD, CEng
Andy Pearson is managing director of contracts at Star Refrigeration, Ltd., Glasgow, United Kingdom.
Jon Evans, Engineer, Sustainable Engineering Group, Madison, WI: Corrective action and/or policy changes in Chicago fire departments related to fire response; changes in building codes?
Andy Pearson: Other engineering disasters prompted change in codes, standards, and regulations, for example the 1905 explosion of the Grover Shoe factory in Brockton, MA. This resulted directly in the publication of the ASME power boiler code ten years later. However, as far as I am aware, the Cold Storage Fire did not result in any change to the fire department's operational procedures, habit, or practice, although the event had a profound effect on the staff. It was, at that time, and remains to this day, the greatest single loss of life in a fire in the City of Chicago. The fire department erected a memorial at Oak Woods Cemetery and also relocated the copper statue of Christopher Columbus that had stood at the entrance to the cold storage building to the fire station manned by 51 Engine Company on South 63rd and Wentworth.
Peggy Niebergall (Member):  5/18/2010 1:46 PM
I was wondering if there were any records kept of those who bought the ice making machines. I understand my great-grandfather bought an ice making machine at the fair and brought it to Wheeling, West Virginia for Niebergall's Ice Company.