- Email: [email protected]
Fire disaster following LPG tanker explosion at Chala in Kannur (Kerala, India): August 27, 2012
burns 39 (2013) 1479–1487
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/burns
Fire disaster following LPG tanker explosion at Chala in Kannur (Kerala, India): August 27, 2012 Pramod Kumar * Burns and Plastic Surgery, Kasturba Medical College, Manipal 576 104, Karnataka, India
article info
abstract
Article history:
A fire disaster following LPG tanker explosion occurred at Chala bypass, Kannur, Kerala,
Accepted 2 April 2013
India on August 27, 2012. The three chambered tanker with total 16 tonnes (162.57 quintal) LPG collided with a road divider and exploded thrice. A total of 41 people became victims
Keywords:
during first blast; out of which 20 died in various hospitals. Five people remained inside the
LPG
house after first blast and escaped unhurt from the zone of accident before second blast. All
Disaster
the victims were transferred to various hospitals; of these, six were transferred to the burns
Mass burn casualty
unit of the Kasturba Hospital, Manipal (320 km from Chala). Five (5/6) were transferred
LPG tanker burst
within 1–5 days at our burns unit suffered 31–72% total body surface area (TBSA) burn, none
Clinical forensic
had external injuries. One (1/6) was transferred on 20th day as a follow up case of 15% TBSA
Limited access dressing
burn with 4% residual raw area and diabetes mellitus. Except one, all were managed conservatively using Limited access dressings (LAD; Negative Pressure Wound Therapy). One of the patient wound bed prepared under LAD and on 41 post burn day underwent split skin grafting under LAD. Out of the six patients admitted at the burns unit, two (2/6) admitted patients expired (one due to inhalation injury and another due to sepsis with multiple organ failure). One survivor (1/4) developed sepsis related liver dysfunction with hepatomegaly but recovered well. The total hospital stay of survivors at the burns unit varied from 8 to 60 days (mean hospital stay 36.5 days). All the victims who developed psychological symptoms were treated by psychiatrists and counselled before discharge. Three of survivors developed psychological symptoms. Two of them (2/3) developed mixed anxiety-depression disorder (ICD 10 code F41.8) and one of these two showed grief reaction too (ICD 10 code F43.23). One victim (1/3) developed non-organic insomnia (ICD 10 code F51.0) and responded to counselling. The article describes the incident, mechanism of the incident, injuries sustained, author, explanations on pattern of burn and suggestions in relation to future safety measures. # 2013 Published by Elsevier Ltd and ISBI.
1.
Introduction
Disaster is a crisis situation that far exceeds our capabilities to recover. In other words, a disaster is defined as a level of disruption which cannot be absorbed by the adjustment capacity of the affected community within its resources. Usually the resources for health management are limited,
overstretched and in disaster it is disrupted. Many people develop psychological symptoms following disasters, and yet the vast majority of them recover due to the resilience of human nature. At the same time, many of those people will develop psychological disorders such as major depression, generalized anxiety, and posttraumatic stress disorder. Many more experience non-specific distress, somatic complaints and other medical health conditions. For better medical and
* Tel.: +91 820 2571201. E-mail addresses: [email protected], [email protected]. 0305-4179/$36.00 # 2013 Published by Elsevier Ltd and ISBI. http://dx.doi.org/10.1016/j.burns.2013.04.004
1480
burns 39 (2013) 1479–1487
psychological health management it is advisable to have better plan and better preparedness. This strategy for preparedness, prevention and minimizing risk of disaster is better formulated in pre-disaster phase based on previous experiences. A fire disaster following LPG tanker explosion occurred at Chala bypass, Kannur, Kerala, India on August 27, 2012. The present fire disaster is being recorded for future reference, improvement in similar disaster prevention plan and better planned care in future.
2.
Methods
The information was collected by the author from media, victims, eye witnesses and rescuers. Also, the author visited the site after 17 days of the accident and collected the information. The District Collector and District Medical Officer were interviewed. Author visited Pariyaram Medical College where he met three victims waiting for discharge. Author also studied the various photographs taken subsequent to accident (provided by the relatives of the victims).
3.
Observations
3.1.
Pre-disaster warning
The driver of the LPG tanker quickly informed about the incidence to the local people and advised not to light fire, use mobiles or to manipulate electric switches. Soon local authorities were informed and electric supply was cut. The cause of fire could not be found. Though local people in unorganized way started running and informing people, there was no continuous announcement for what to do/not to do.
3.2.
The disaster
3.2.1.
The event
The disaster occurred on August 27, 2012 at Chlala village in Kannur district of Kerala state of India (Fig. 1a and b). An Indian Oil Corporation LPG tanker rammed a divider in an attempt to overtake a vehicle at the Chala bypass near Bhagavathy Temple at 11 p.m. The tanker containing LPG had three chambers with total 16 tonnes (162.57 quintal) LPG in it. The wall of the container was made of 1 cm thick steel. The LPG vapour was seen to flow slowly all around at the ground level like cloud of white colour gas (as described by eye witness) that burst into flames within 20 min. The blast sound was heard up to 25 km distance and was soon followed by two more blasts at 3 min interval. People witnessed sudden short duration fire rising upwards above the coconut tree (approximately 50 feet high) within involved area. After third blast approx one third part of container was thrown like missile approximately 60–70 feet above the ground passing over coconut trees (Eye witness) and later found about 500 m away from the site of accident in a field.
3.2.2. The impact 3.2.2.1. On immediate medical need. All the coconut trees in the zone of fire were burnt. In the centre of the involved zone
all the burnt green vegetation included more than 50 feet high coconut trees up to the top and grass at the ground level, but those towards the periphery the height of burnt portion of coconut trees was seen to reduce progressively; green leaves in upper portion of trees were spared (site marked A in Fig. 2). Gross at ground level at the periphery of the affected zone (site marked as B in Fig. 2) was not burnt. The vegetations on the other side of seven feet high brick wall were also spared (Fig. 3). There were total 41 victims; out of which 20 died in various hospitals. Five people (Three males – 40 year, 55 year, 65 year and two females – 48 year, 55 year) moved away from the zone of accident after leak but before it burst into fire and escaped. Five people (two males – 13 year, 27 year; three females – 24 year, 50 year and 93 year) could not escape before first blast and remained inside the house and then could escape unhurt from the zone of accident before second blast. One of the victims who was disabled due to pre-existing illness less than 50 m from the site of accident could not escape and succumbed to his extensive injuries. At the time of first blast all those who were curious and out of their house were burnt severely. Those who were watching from inside the house with doors open sustained mild to moderate burn.
3.2.2.2. On physical environment. Total 7 shops, thirty five individual houses were burnt. Three houses within 50 feet from the site of accident were completely burnt. Two thatched roofs were blown off. Walls of the houses and windows were damaged from outside. There were no significant damages inside the houses at the peripheral zone. Total 60 vehicles were reported to be damaged. There was no damage to roads and communication system. Electric supply was disrupted for two days. 3.2.2.3. On veterinary population. Within the zone of accident two dogs, two hens and one cat were found dead. 3.2.2.4. Psychological consequences. Initially all the victims showed fear, anxiety and helplessness. Next day some of them showed frustration and anger (expressed on the people taking photograph of the site of accident). Within next few days following sharing of the experiences, emotional support (by friends, relatives and local psychologists) and social support for disruption (by Government and LPG company) losses were accepted and process of rebuilding life started. All the patients were assessed using self-reported and observer rated questionnaires and interviews [1]. Three of survivors developed psychological symptoms. Two of them developed mixed anxiety-depression disorder (2013 ICD-10-CM Diagnosis Code F41.8) and one of these two showed grief (adjustment) reaction too (2013 ICD-10-CM Diagnosis Code F43.23). One victim developed non-organic insomnia (ICD-10: version 2010: Code F51.0) and responded to counselling but had anxiety related distress even after 1 month of discharge while all other had started normal life. 3.3.
All the victims were transferred to various hospitals
Total seven victims were shifted to adjacent state Karnataka, out of which six victims (Table 1) were transferred to the burns unit of the Kasturba Hospital, Manipal (320 km from Chala). Of
burns 39 (2013) 1479–1487
1481
Fig. 1 – (a) Arial view (sketch) of the area involved in ‘Chala’ LPG fire disaster. (A–H) represents location of houses where all 20 deaths occurred due to disaster. House A and B were situated against 7 feet high wall between 150 and 200 m towards north from the site of accident. House F was situated in front of another house H (on 10 feet high land piece) with 3 feet high boundary. LPG vapour on south side of the accident was seen flowing mainly along the canal represented by dark line in the sketch. (b) Diagram (cross section) showing slope on either side of the road at the site of accident. Involved area extended up to 300 m on the steeper side towards south of the road as compared to that on the north side.
these, six victim, five were transferred within 1–5 days at our burns unit suffered 31–72% total body surface area (TBSA) burn, none had external injuries. One was transferred on 20th day as a follow up case of 15% TBSA burn with 4% residual raw area and diabetes mellitus. Out of the six patients admitted at our burns unit, two expired (Table 1) (one due to inhalation injury and another due to sepsis with multiple organ failure).
3.4.
The site of accident and extent of injuries
The site of the accident was situated on the busy bypass road with slope on either side or the road (Fig. 1b). On one side of the road the slope was steeper and had approx 4 feet (width) 4 feet (depth) canal (Fig. 1a). There were few shops and sparse houses on either side of the road. The accident involved area between 50 m distance on one side of the accident to 100 m on
the other side along the road. The difference in the distance on same level road could be explained by the direction of wind and force of LPG leak and blast. The accident involved area between 200 m on north side of the road and 300 m on south side from the point of the accident (Fig. 1a). The difference in the distance involved on either side of the road could be explained by extent of slope- steeper side of the road was on the south side and evidenced flow of LPG vapour under influence of gravitational force for a greater distance) (300 m). Two feet thick cloud like layer of LPG vapour was seen flowing slowly of the ground before the first fire burst mainly along a canal on steeper south side of the road (Eye witness). No effect of accident was seen on other side of a seven feet brick wall situated at 50–60 m away from the site of accident. Also, no damage was seen in a house situated 100 m away from the site at approximately 10 feet high land piece and surrounded by 3
1482
burns 39 (2013) 1479–1487
Fig. 2 – Photograph (taken on 17 day from the centre of disaster hit area after taking position near canal on the south side) showing involvement of the coconut tree up to greater height towards centre (A) of the area as compared to that at the periphery (B) where top leaves were spared. Also at the periphery gross was green but near centre gross was brown.
feet high brick wall (Fig. 1a). Between 150 m to 200 m on north side of the accident spot, six people were standing between their houses and seven feet high brick wall (House A and B in Fig. 1a). All six developed extensive burn (70 to 100% TBSA burn, Mean 83.67%). One victim was inside the house near the door developed 42% TBSA burn. Mean TBSA burn of all seven
Fig. 3 – Photograph (taken from position near house A shown in Fig. 1a) showing no effect of disaster on vegetation on the other side of a seven feet high brick wall.
cases in this area was 72.71%. One victim, an old case of paraparesis, was within 50 m on north side sustained 35% TBSA burn and expired (? inhalation injury) as he could not escape. On the south side at about 100 m from the site of accident, eight residents in two houses situated in open ground (house D and E in Fig. 1a) became victims and expired with Mean TBSA burn 68.13 (range 45–90%). Two victims who were standing close to a house (house G in Fig. 1a) with 3 feet high boundary wall situated on a 10 feet high land piece (reflecting hot gases and flame) sustained 90% TBSA burn each and later succumb
Table 1 – Summary of the victims treated at Burns Unit of Kasturba Medical College, Manipal. Case no
Post burn day of admission
1
<1 day
52
Male
72%
2
<1 day
42
Female
41% with sepsis related jaundice with anxiety depressive disorder
3
3 days
45
Female
55% with inhalation injury
4
4 days
16
Male
31% with distressed sleep
5
4 days
55
Male
6
20 days
60
Male
33% with anxiety depressive disorder and grief reaction 15% burn with 4% raw area with Diabetes Mellitus
Age (yr)
Sex
Percent TBSA burn
Ares involved Whole body except circumferential area around waist, upper buttock and scalp Both leg and foot circumferential, posterior thigh patchy (right > left), upper back, back of neck, face, both upper extremities (right > left), patchy burn upper abdomen Circumferential burn both leg and foot, upper back, back of neck, face, both upper extremities, patchy burn chest and upper abdomen, right buttock) Both feet, face, patches over abdomen, back, upper buttock, both upper extremities Both lower extremities circumferential up to mid-thigh, both hands (right > left), face and neck Both lower extremities (right > left)
Treatment given/outcome Conservative (NPWT using LAD)/ Expired on 5th day (Sepsis with MOF) NPWT using LAD for wound bed preparation and granulation grafting under LAD/Survived (Hospital days 60)
Conservative (NPWT using LAD) (ventilator support immediately after admission/Expired on 6th day (Sudden death) Conservative (NPWT using LAD)/ Survived (Hospital days 29) Conservative (NPWT using LAD)/ Survived (Hospital days 29)
Conservative (NPWT using LAD)/ Survived (Hospital days 8 in burns unit + 20 days in surgical unit outside)
burns 39 (2013) 1479–1487
to their injuries. Near house G (in Fig. 1a) one victim sustained 55% TBSA burn with respiratory burn and expired on 6th day.
3.5.
done to prevent communicable diseases. Debris from the site of accident on the road was cleared after police and legal formalities.
Resources 3.8.
Police chief and DMO reached on the spot within half an hour. Well placed disaster management plan was activated. Under direct supervision of Kannur District Collector, District Medical Officer, and Deputy District Education Media Officer-DMO Health a local team was formed that mainly people from a nearby village Edakadu. The prominent members of the team were: President, Edakadu Gram Panchayat, Medical District Education and Media Officer (DMO-Health) from Kannur, Health Inspectors from Edakadu Primary Health Centre. One hundred and fifty personnel from various fields (administration, Police, Fire force, residents from adjacent area) who were familiar to the place and people were required to manage the disaster. Initially victims were shifted to various hospitals by local people in their private or hired vehicle. Within 30 min 10 ambulances (including four High tech ambulances with facilities to transport critically ill patient hired from neighbouring district) were arranged to transport the victims. Initially victims were shifted to nearby hospitals where triage was done after first aid and then shifted to tertiary care hospitals in adjacent district and state.
3.6.
Difficulties faced
Problems in medical management of victims were experienced due to lack of high tech trauma centre, burns unit and ambulances in the state. It was difficult to manage crowd of volunteers gathered at the site and was managed by restricting their number. Local people objected to making videos of sufferings/ damages and uploading to social networking site immediately after the accident. As the people involved in rescue work initially from local known persons, no incidence by unscrupulous people was reported.
3.7.
1483
Rescue and relief work
Since communication and transport were not disrupted, rescue and relief work was relatively easy. Blast sound triggered immediate post-disaster response. Road was temporarily blocked and traffic were diverted to other routes. Availability of District Collector, District Medical Officer, and Police chief along with other officials on the site within 30 min for supervising and directing relief work made decision making very quick. Since the president of adjacent Gram Panchayat, was familiar with the place, people and custom of that area he was appointed as chief co-ordinator. Permanent District Information Centre and Temporary information centre set up by police served as information centre. Relief activities, local issues and victims condition were updated daily through mass media. Keeping in mind the available resources in the hospitals, victims were evenly distributed to all hospitals and clinics. All dead animals were removed from the area for safe disposal as per local predefined procedure. Phenol and Bleaching Powder were sprinkled; Water chlorination was
Social work
Appropriate compensation was declared from government and Indian Oil Corporation to minimize post disaster psychological problems, social disharmony and as a reassurance measure for the grieved relatives and victims.
4.
Mitigation measures
Following measures were taken to minimize risk in future:
(a) Divider was removed and road modifications done where ever suggested by competent authorities. (b) Security enhanced and traffic rules and regulations were strictly reinforced especially for vehicles carrying inflammable materials. (c) Emphasis was made to enforce standardized communication protocol, and safety protocol. (d) Rules related to essential drugs and supplies of emergencies were reviewed and strictly enforced. (e) Revamping of own government run burns and trauma centre along with high tech ambulances with critical care facilities and life support systems. (f) Liaison with external non-government and government agencies (administrative, health sector, relief related) were reviewed. (g) Exhaustive training and hi tech equipment to fire, rescue and police personnel to be given to deal with such disasters. (h) Education: timely cutting of electricity, Stopping use of mobile to avoid burst of LPG into flames.
5.
Discussion
Scientific literature on large scale LPG disaster is scanty in English literature. Majority of such incidences have been recorded in daily newspapers. Similar LPG burst has been mentioned in daily national newspaper (The Hindu, September 3, 2012, http://www.thehindu.com/news/states/kerala/article3854053.ece). The report mentions that on December 31, 2009, when an LPG tanker rammed a car at the Punthentheruvu junction at 3.50 a.m. and burst into flames. Twelve of the 18 persons who suffered burns died, most of them natives of Karunagapally and nearby areas who rushed in for rescue operations. Those killed were from the economically weaker sections, many of them daily wage earners. The tanker was carrying 18 tonnes of LPG from Kochi to the IOC’s refilling plant at Paripally. A major disaster was averted in Kochi on September 10, 2012 after 16 tonnes of cooking gas was accidentally released into the open at the Udayamperoor bottling plant of the Indian Oil Corporation (IOC), Kochi, Kerala. The Hindu, September 11, 2012 (http://www.thehindu.com/news/cities/
1484
burns 39 (2013) 1479–1487
Kochi/article3884493.ece) and The Times of India, September 11, 2012 (http://articles.timesofindia.indiatimes.com/201209-11/kochi/33762205_1_tanker-drivers-gas-leak-lpg). The accident occurred around 4.30 p.m. when the driver of a tanker moved the vehicle which was connected to the plant in a bay for transfer of gas. The valve broke, leaving no scope for plugging the leak. The entire quantity of gas in the tanker was released. The measures taken to avert a major fire accident due to LPG leak were: Stopping the movement of traffic for nearly an hour, cutting of electricity supplies. Residents in the vicinity of the bottling plant were alerted about the gas leak immediately and advised to leave their houses for safer places. In the tragedy of San Juanico (20 km North of Mexico City), on November 19, 1984 the disaster started due to LPG leakage, probably a pipe leakage or rupture due to excess pressure the storage tanks contained 11,000 m3 of a mixture of propane and butane (Pemex State Oil Company). The vapour cloud was ignited around 5:40 a.m. and was followed by an extensive fire at the plant area. The first explosion was registered on the seismograph at the University of Mexico at 05 h 44 min 52 s and was followed by a dozen explosions within the next hour, some of them of BLEVE type (Boiling Liquid Expanding Vapour Explosion) due to rupture of one or more storage tanks. Two of the explosions had an intensity of 0.5 on the Richter scale. Unburned and burning gas entered the houses south of the plant area and set fire to everything. Blast waves from the explosions not only destroyed a number of houses but also shifted several cylindrical tanks from their supports and added more gas to the fire. The smaller spheres and some of the cylinders exploded and fragments and even whole cylinders weighing around 30 tonnes were scattered over distances ranging from a few to up to 1200 m [2]. In the disaster reported in the present study the accident occurred when an Indian Oil Corporation LPG tanker rammed a road divider in an attempt to overtake a vehicle. The tanker containing LPG had three chambers (steel wall of 1 cm thickness) with total 16 tonnes (162.57 quintal) LPG in it. The LPG vapour was burst into flames within 20 min. The blast sound was heard up to 25 km distance and was soon followed by two more blasts at 3 min interval. After third blast approx one third part of container was thrown like missile 60–70 feet above the ground over coconut trees and later found about 500 m away from the site of accident in a field. LPG was seen by eyewitnesses flowing along the ground surface as LPG is heavier than atmospheric gases. The direction and speed of flow was determined by the direction of the wind and force of the burst. The coconut trees in the area were burnt up to top near the centre of the involved area, but the height of heat damaged portion of the trees gradually reduced at the periphery (Fig. 2 from Position A to more peripheral position B) indicating upper limit of heat making a shape of a dome. Since leaked LPG burns quickly and light weight heated gas expands over a wider area and moves upwards. Hence rapidly expanding burning LPG takes curved path from centre to periphery with convexity towards ground. Upward moving expanding hot gas contacts ground for shorter duration at the periphery (i.e. moves up towards periphery making a curve with convexity towards ground). This may be the possible explanation for relatively green vegetation at the periphery
(Fig. 2 at position B). Maximum death were recorded when victim positioned near seven feet high wall or near high house acting as block to reflect the blast wave and flow of hot air along with flame (Fig. 1a). Vegetation at the ground level at the periphery of affected zone showed little effect of heat suggesting that the lighter hot air and flame takes upward curve when force of blast wave becomes weak. Liquefied petroleum gas (LPG), which is used as a type of fuel, is stored as a liquid under high pressure in tanks. Immediate and sudden explosion of these tanks can release a large amount of gas and energy into the environment and can result in serious burns. Bozkurt et al. [3] have reported 18 patients injured due to LPG burns in five car LPG tank explosions incidents. The reported inhalation injury in 11 cases with varying degrees of severity, and 7 patients subsequently required mechanical ventilation. The explosions resulted from weakening of the tank wall (n = 2), crash impact (n = 2), and gas leakage from the tank (n = 1). In the LPG tanker explosion reported in the present study only one case of the six admitted in our burns centre had inhalation injury and required immediate ventilator support. The open space and distance of the victims could be the possible explanation for this difference in incidence of the inhalation injury.
5.1.
The pattern of burn (clinical forensic)
Based on the study of the victims in the field and admitted patient, the author found that the following points may be helpful to clinical forensic experts to further develop the clinical forensic in LPG burst burn (Fig. 4):
Fig. 4 – Pattern/type of burn at various positions in an open disaster area – Position A: extensive deep burn; sole burn if steps on the hot ground; Position B: extensive burn on the side of the body facing blast side and relatively deeper burn on lower positioned part of the body as LPG flows at the ground level; Position C: extensive but relatively superficial burn with deeper burn on the lower positioned part of the body facing the side of blast; Position D: less extensive mainly superficial burn on the part of body facing the side of blast; hot gases being relatively lighter starts moving upwards once the movement due to blast force become weak at some distance from the blast site and spares the relatively lower placed part of the body. Note: Sole burn may occur if patient steps on hot objects (thrown like missiles). As the flame disappears rapidly like flash, body parts covered by cotton cloth are usually spared. Respiratory burn may occur if victim’s position is near blast site or LPG is inhaled before catching fire.
burns 39 (2013) 1479–1487
1485
Fig. 5 – Patient (Case 3 in Table 1) at position A of Fig. 4. Her back was involved extensively.
(A) Victims near the site of accident in open area (if not protected by cloth or wall) develop extensive whole body burn. Burn is relatively deeper over lower positioned parts of the body (i.e. legs. hands, forearm) (Fig. 5). Pattern of depends on the type and material of cloth. Pattern is also determined by body posture and attempt to escape from the site. (B) Victims at the periphery of the fire zone (a) Mainly one side of burn indicates patient was exposed to fire for the short duration and probably he was positioned more towards the periphery. Victim’s position and posture may be determined by examining pattern of burn as more burnt surface of the body faces towards blast side (e.g. back involvement is more if victim is running away from the blast site, rt side body is involved more if victim tries to take right turn in an attempt to escape). (b) Foot burn is superficial/there will be no foot burn if the victim is positioned at the periphery in open area (Fig. 6). (c) At the extreme periphery (in open area) as the hot gases and fumes moves very fast towards upper strata, face, neck and exposed part of the body sustain superficial burn. Higher the position of the body part, more superficial burn it sustains (Fig. 7). (d) If victim is near a high wall even at the periphery of the affected zone, he/she may get extensive burn as expanding fires and hot gases gets reflected from the
Fig. 7 – Pattern of burn in a victim (Case 4 in Table 1) between position C & D of Fig. 4. Patient also had feet burns. Wounds were managed under LAD.
wall and produce turbulence in flow, in turn turbulence may induce movement of gases towards ground. The duration of exposure will also be more due to turbulence as well as due to failure of gases to spread to wider area (i.e. accumulation of more gas) due to blockade by the wall (Fig. 8). (C) Sole bun indicates walking on the hot ground near the site of accident or walking on the hot missile like objects due to blast. One of the victims the author examined at Periyaram Medical College had sole burn. History revealed that at the time of fire he was in his shop within 5 m from the site of the blast. (D) The significant difference in the level of burn of both lower extremities usually indicates the movement of fire and heat with blast wave (curved path with convexity towards ground) (Fig. 9). Usually the hand (upper extremity) will be involved to varying extent on the side the lower extremity is involved up to or above knee level. (E) Those who inhale gases or very close to blast are likely to develop inhalation injury (case 3 Table 1).
5.2.
Prevention of burn
Various studies have been done to find the model, mechanism and impact of boiling liquid expanding vapour explosion (BLEVE) to formulate the preventive strategies associated with BLEVEs [4–12]. Author would like to suggest following simple measures to avoid or minimize the extent of burn after
Fig. 6 – Patten of burn in a victim (Case 6 in Table 1) at extreme periphery (position D in Fig. 4) taking left turn to run away. Wounds were managed under LAD.
Fig. 8 – Pattern of burn in victim (Case 1 in Table 1) positioned close to a wall peripheral area. Wounds were managed under LAD.
1486
burns 39 (2013) 1479–1487
LAD design has notable advantages like wound isolation that reduces chance of wound colonization and safe disposal of infected materials (important factor to reduce hospitalacquired infections), while avoiding some major disadvantages such as opacity of dressing materials, inaccessible offensive smelling wound environment, and relatively high treatment costs. In LAD a definite intermittent negative pressure regimen is followed. The intermittent negative pressure (cycle of 30 min suction and 3½ h rest) is effective. Overall, the LAD is a safe and effective alternative to conventional dressing methods. LAD is an excellent research tool for wound healing as frequent/ continuous record of wound healing is possible without disturbing the wound healing process. LAD is an effective dressing for limb salvage in cases of acute and chronic complex wounds. Leech effect prevents wound related systematic response syndrome and sepsis.
Conflicts of interest There is no conflict of interest that could inappropriately influence (bias) my work.
Acknowledgements
Fig. 9 – Pattern of burn in a victim at position C of Fig. 4.
accident apart from measure taken at technology level and administrative level:
(a) Position behind the wall opposite side of accident will protect the body from exposure to heat. (b) One should move in open space away from the accident site to avoid getting trapped in turbulence due to reflected blast waves by obstructions (like wall) causing more extensive circumferential burn. (c) Covering the body with ordinary cotton sheet (and wearing cotton socks with pants tucked in it) while escaping may minimize the extent of burn. (d) One should leave the houses immediately for safer locations.
5.3. Use of limited access dressing (LAD) [13] during disaster Limited access dressing was found useful during disaster for managing large wounds. Limited access dressing (LAD) is a moist wound dressing with negative pressure. This is economical dressing that does not require frequent changing and saves resources and time. It provides limited access to the wound through two small ports for both dressers and pathogens. The
The author acknowledges the input by all eye witness (especially Mr. Bhaskaran P.K.) and support by Rajan Kelkar IAS (District Collector), Dr. R. Ramesh (District Medical Officer), Mr. K.N. Ajay (Medical District Education and Media Officer, DMO-Health, Kannur), Mr. C.H. Rajeevan (Health Inspector, Edakadu Primary Health Centre), Mr. Anish Babu (Health Inspector, Edakadu Primary Health Centre), Mr. K.V. Raveendran (President, Edakadu Gram Panchayat – Edakadu is the name of adjacent village), and administration of Pariyaram Medical College without which the study would not have been possible. The author also acknowledges the inputs by victims (especially Mr. Bhaskaran P.K., Mr. Kunhi Krishnan, Mr. Vineet Krishnan, and Mr. Kunhi Kanan) and relatives of the victims (especially Mr. A.K. Surendran) for extending all the help and providing photographs collected by them.
references
[1] Gupta A, Deepika S, Taly AB, Srivastava A, Surender V, Thyloth M. Quality of life and psychological problems in patients undergoing neurological rehabilitation. Ann Indian Acad Neurol 2008;11:225–30. [2] Arturson G. The tragedy of San Juanico – the most severe LPG disaster in history. Burns Incl Therm Inj 1987; 13:87–102. [3] Bozkurt M, Kulahci Y, Zor F, Kapi E. Burn injuries related to liquefied petroleum gas-powered cars. J Burn Care Res 2008;29:897–901. [4] Abbasi T, Abbasi SA. The boiling liquid expanding vapour explosion (BLEVE): mechanism, consequence assessment, management. J Hazard Mater 2007;141:489–519. [5] Ogle RA, Megerle MV, Morrison 3rd DR, Carpenter AR. Explosion caused by flashing liquid in a process vessel. J Hazard Mater 2004;115:133–40.
burns 39 (2013) 1479–1487
[6] Stawczyk J. Experimental evaluation of LPG tank explosion hazards. J Hazard Mater 2003;96:189–200. [7] Gong YW, Lin WS, Gu AZ, Lu XS. A simplified model to predict the thermal response of PLG and its influence on BLEVE. J Hazard Mater 2004;108:21–6. [8] Stawczy KJ. Experimental evaluation of LPG tank explosion hazards. J Hazard Mater 2003;96:189–200. [9] Papazoglou IA, Aneziris ON. Uncertainty quantification in the health consequences of the boiling liquid expanding vapour explosion phenomenon. J Hazard Mater 1999;67(June):217–35. [10] Shebeko YN, Bolodian IA, Filippov VN, Navzenya VY, Kostyuhin AK, Tokarev PM, et al. A study of the behaviour
1487
of a protected vessel containing LPG during pool fire engulfment. J Hazard Mater 2000;77:43–56. [11] Paltrinieri N, Landucci G, Molag M, Bonvicini S, Spadoni G, Cozzani V. Risk reduction in road and rail LPG transportation by passive fire protection. J Hazard Mater 2009;167:332–44. [12] Paltrinieri N, Bonvicini S, Spadoni G, Cozzani V. Costbenefit analysis of passive fire protections in road LPG transportation. Risk Anal 2012;32:200–19. http://dx.doi.org/ 10.1111/j.1539-6924.2011.01654.x [discussion 220–3]. [13] Kumar P. Exploiting potency of negative pressure in wound dressing using limited access dressing and suction-assisted dressing. Indian J Plast Surg 2012;45:302–15.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/burns
Fire disaster following LPG tanker explosion at Chala in Kannur (Kerala, India): August 27, 2012 Pramod Kumar * Burns and Plastic Surgery, Kasturba Medical College, Manipal 576 104, Karnataka, India
article info
abstract
Article history:
A fire disaster following LPG tanker explosion occurred at Chala bypass, Kannur, Kerala,
Accepted 2 April 2013
India on August 27, 2012. The three chambered tanker with total 16 tonnes (162.57 quintal) LPG collided with a road divider and exploded thrice. A total of 41 people became victims
Keywords:
during first blast; out of which 20 died in various hospitals. Five people remained inside the
LPG
house after first blast and escaped unhurt from the zone of accident before second blast. All
Disaster
the victims were transferred to various hospitals; of these, six were transferred to the burns
Mass burn casualty
unit of the Kasturba Hospital, Manipal (320 km from Chala). Five (5/6) were transferred
LPG tanker burst
within 1–5 days at our burns unit suffered 31–72% total body surface area (TBSA) burn, none
Clinical forensic
had external injuries. One (1/6) was transferred on 20th day as a follow up case of 15% TBSA
Limited access dressing
burn with 4% residual raw area and diabetes mellitus. Except one, all were managed conservatively using Limited access dressings (LAD; Negative Pressure Wound Therapy). One of the patient wound bed prepared under LAD and on 41 post burn day underwent split skin grafting under LAD. Out of the six patients admitted at the burns unit, two (2/6) admitted patients expired (one due to inhalation injury and another due to sepsis with multiple organ failure). One survivor (1/4) developed sepsis related liver dysfunction with hepatomegaly but recovered well. The total hospital stay of survivors at the burns unit varied from 8 to 60 days (mean hospital stay 36.5 days). All the victims who developed psychological symptoms were treated by psychiatrists and counselled before discharge. Three of survivors developed psychological symptoms. Two of them (2/3) developed mixed anxiety-depression disorder (ICD 10 code F41.8) and one of these two showed grief reaction too (ICD 10 code F43.23). One victim (1/3) developed non-organic insomnia (ICD 10 code F51.0) and responded to counselling. The article describes the incident, mechanism of the incident, injuries sustained, author, explanations on pattern of burn and suggestions in relation to future safety measures. # 2013 Published by Elsevier Ltd and ISBI.
1.
Introduction
Disaster is a crisis situation that far exceeds our capabilities to recover. In other words, a disaster is defined as a level of disruption which cannot be absorbed by the adjustment capacity of the affected community within its resources. Usually the resources for health management are limited,
overstretched and in disaster it is disrupted. Many people develop psychological symptoms following disasters, and yet the vast majority of them recover due to the resilience of human nature. At the same time, many of those people will develop psychological disorders such as major depression, generalized anxiety, and posttraumatic stress disorder. Many more experience non-specific distress, somatic complaints and other medical health conditions. For better medical and
* Tel.: +91 820 2571201. E-mail addresses: [email protected], [email protected]. 0305-4179/$36.00 # 2013 Published by Elsevier Ltd and ISBI. http://dx.doi.org/10.1016/j.burns.2013.04.004
1480
burns 39 (2013) 1479–1487
psychological health management it is advisable to have better plan and better preparedness. This strategy for preparedness, prevention and minimizing risk of disaster is better formulated in pre-disaster phase based on previous experiences. A fire disaster following LPG tanker explosion occurred at Chala bypass, Kannur, Kerala, India on August 27, 2012. The present fire disaster is being recorded for future reference, improvement in similar disaster prevention plan and better planned care in future.
2.
Methods
The information was collected by the author from media, victims, eye witnesses and rescuers. Also, the author visited the site after 17 days of the accident and collected the information. The District Collector and District Medical Officer were interviewed. Author visited Pariyaram Medical College where he met three victims waiting for discharge. Author also studied the various photographs taken subsequent to accident (provided by the relatives of the victims).
3.
Observations
3.1.
Pre-disaster warning
The driver of the LPG tanker quickly informed about the incidence to the local people and advised not to light fire, use mobiles or to manipulate electric switches. Soon local authorities were informed and electric supply was cut. The cause of fire could not be found. Though local people in unorganized way started running and informing people, there was no continuous announcement for what to do/not to do.
3.2.
The disaster
3.2.1.
The event
The disaster occurred on August 27, 2012 at Chlala village in Kannur district of Kerala state of India (Fig. 1a and b). An Indian Oil Corporation LPG tanker rammed a divider in an attempt to overtake a vehicle at the Chala bypass near Bhagavathy Temple at 11 p.m. The tanker containing LPG had three chambers with total 16 tonnes (162.57 quintal) LPG in it. The wall of the container was made of 1 cm thick steel. The LPG vapour was seen to flow slowly all around at the ground level like cloud of white colour gas (as described by eye witness) that burst into flames within 20 min. The blast sound was heard up to 25 km distance and was soon followed by two more blasts at 3 min interval. People witnessed sudden short duration fire rising upwards above the coconut tree (approximately 50 feet high) within involved area. After third blast approx one third part of container was thrown like missile approximately 60–70 feet above the ground passing over coconut trees (Eye witness) and later found about 500 m away from the site of accident in a field.
3.2.2. The impact 3.2.2.1. On immediate medical need. All the coconut trees in the zone of fire were burnt. In the centre of the involved zone
all the burnt green vegetation included more than 50 feet high coconut trees up to the top and grass at the ground level, but those towards the periphery the height of burnt portion of coconut trees was seen to reduce progressively; green leaves in upper portion of trees were spared (site marked A in Fig. 2). Gross at ground level at the periphery of the affected zone (site marked as B in Fig. 2) was not burnt. The vegetations on the other side of seven feet high brick wall were also spared (Fig. 3). There were total 41 victims; out of which 20 died in various hospitals. Five people (Three males – 40 year, 55 year, 65 year and two females – 48 year, 55 year) moved away from the zone of accident after leak but before it burst into fire and escaped. Five people (two males – 13 year, 27 year; three females – 24 year, 50 year and 93 year) could not escape before first blast and remained inside the house and then could escape unhurt from the zone of accident before second blast. One of the victims who was disabled due to pre-existing illness less than 50 m from the site of accident could not escape and succumbed to his extensive injuries. At the time of first blast all those who were curious and out of their house were burnt severely. Those who were watching from inside the house with doors open sustained mild to moderate burn.
3.2.2.2. On physical environment. Total 7 shops, thirty five individual houses were burnt. Three houses within 50 feet from the site of accident were completely burnt. Two thatched roofs were blown off. Walls of the houses and windows were damaged from outside. There were no significant damages inside the houses at the peripheral zone. Total 60 vehicles were reported to be damaged. There was no damage to roads and communication system. Electric supply was disrupted for two days. 3.2.2.3. On veterinary population. Within the zone of accident two dogs, two hens and one cat were found dead. 3.2.2.4. Psychological consequences. Initially all the victims showed fear, anxiety and helplessness. Next day some of them showed frustration and anger (expressed on the people taking photograph of the site of accident). Within next few days following sharing of the experiences, emotional support (by friends, relatives and local psychologists) and social support for disruption (by Government and LPG company) losses were accepted and process of rebuilding life started. All the patients were assessed using self-reported and observer rated questionnaires and interviews [1]. Three of survivors developed psychological symptoms. Two of them developed mixed anxiety-depression disorder (2013 ICD-10-CM Diagnosis Code F41.8) and one of these two showed grief (adjustment) reaction too (2013 ICD-10-CM Diagnosis Code F43.23). One victim developed non-organic insomnia (ICD-10: version 2010: Code F51.0) and responded to counselling but had anxiety related distress even after 1 month of discharge while all other had started normal life. 3.3.
All the victims were transferred to various hospitals
Total seven victims were shifted to adjacent state Karnataka, out of which six victims (Table 1) were transferred to the burns unit of the Kasturba Hospital, Manipal (320 km from Chala). Of
burns 39 (2013) 1479–1487
1481
Fig. 1 – (a) Arial view (sketch) of the area involved in ‘Chala’ LPG fire disaster. (A–H) represents location of houses where all 20 deaths occurred due to disaster. House A and B were situated against 7 feet high wall between 150 and 200 m towards north from the site of accident. House F was situated in front of another house H (on 10 feet high land piece) with 3 feet high boundary. LPG vapour on south side of the accident was seen flowing mainly along the canal represented by dark line in the sketch. (b) Diagram (cross section) showing slope on either side of the road at the site of accident. Involved area extended up to 300 m on the steeper side towards south of the road as compared to that on the north side.
these, six victim, five were transferred within 1–5 days at our burns unit suffered 31–72% total body surface area (TBSA) burn, none had external injuries. One was transferred on 20th day as a follow up case of 15% TBSA burn with 4% residual raw area and diabetes mellitus. Out of the six patients admitted at our burns unit, two expired (Table 1) (one due to inhalation injury and another due to sepsis with multiple organ failure).
3.4.
The site of accident and extent of injuries
The site of the accident was situated on the busy bypass road with slope on either side or the road (Fig. 1b). On one side of the road the slope was steeper and had approx 4 feet (width) 4 feet (depth) canal (Fig. 1a). There were few shops and sparse houses on either side of the road. The accident involved area between 50 m distance on one side of the accident to 100 m on
the other side along the road. The difference in the distance on same level road could be explained by the direction of wind and force of LPG leak and blast. The accident involved area between 200 m on north side of the road and 300 m on south side from the point of the accident (Fig. 1a). The difference in the distance involved on either side of the road could be explained by extent of slope- steeper side of the road was on the south side and evidenced flow of LPG vapour under influence of gravitational force for a greater distance) (300 m). Two feet thick cloud like layer of LPG vapour was seen flowing slowly of the ground before the first fire burst mainly along a canal on steeper south side of the road (Eye witness). No effect of accident was seen on other side of a seven feet brick wall situated at 50–60 m away from the site of accident. Also, no damage was seen in a house situated 100 m away from the site at approximately 10 feet high land piece and surrounded by 3
1482
burns 39 (2013) 1479–1487
Fig. 2 – Photograph (taken on 17 day from the centre of disaster hit area after taking position near canal on the south side) showing involvement of the coconut tree up to greater height towards centre (A) of the area as compared to that at the periphery (B) where top leaves were spared. Also at the periphery gross was green but near centre gross was brown.
feet high brick wall (Fig. 1a). Between 150 m to 200 m on north side of the accident spot, six people were standing between their houses and seven feet high brick wall (House A and B in Fig. 1a). All six developed extensive burn (70 to 100% TBSA burn, Mean 83.67%). One victim was inside the house near the door developed 42% TBSA burn. Mean TBSA burn of all seven
Fig. 3 – Photograph (taken from position near house A shown in Fig. 1a) showing no effect of disaster on vegetation on the other side of a seven feet high brick wall.
cases in this area was 72.71%. One victim, an old case of paraparesis, was within 50 m on north side sustained 35% TBSA burn and expired (? inhalation injury) as he could not escape. On the south side at about 100 m from the site of accident, eight residents in two houses situated in open ground (house D and E in Fig. 1a) became victims and expired with Mean TBSA burn 68.13 (range 45–90%). Two victims who were standing close to a house (house G in Fig. 1a) with 3 feet high boundary wall situated on a 10 feet high land piece (reflecting hot gases and flame) sustained 90% TBSA burn each and later succumb
Table 1 – Summary of the victims treated at Burns Unit of Kasturba Medical College, Manipal. Case no
Post burn day of admission
1
<1 day
52
Male
72%
2
<1 day
42
Female
41% with sepsis related jaundice with anxiety depressive disorder
3
3 days
45
Female
55% with inhalation injury
4
4 days
16
Male
31% with distressed sleep
5
4 days
55
Male
6
20 days
60
Male
33% with anxiety depressive disorder and grief reaction 15% burn with 4% raw area with Diabetes Mellitus
Age (yr)
Sex
Percent TBSA burn
Ares involved Whole body except circumferential area around waist, upper buttock and scalp Both leg and foot circumferential, posterior thigh patchy (right > left), upper back, back of neck, face, both upper extremities (right > left), patchy burn upper abdomen Circumferential burn both leg and foot, upper back, back of neck, face, both upper extremities, patchy burn chest and upper abdomen, right buttock) Both feet, face, patches over abdomen, back, upper buttock, both upper extremities Both lower extremities circumferential up to mid-thigh, both hands (right > left), face and neck Both lower extremities (right > left)
Treatment given/outcome Conservative (NPWT using LAD)/ Expired on 5th day (Sepsis with MOF) NPWT using LAD for wound bed preparation and granulation grafting under LAD/Survived (Hospital days 60)
Conservative (NPWT using LAD) (ventilator support immediately after admission/Expired on 6th day (Sudden death) Conservative (NPWT using LAD)/ Survived (Hospital days 29) Conservative (NPWT using LAD)/ Survived (Hospital days 29)
Conservative (NPWT using LAD)/ Survived (Hospital days 8 in burns unit + 20 days in surgical unit outside)
burns 39 (2013) 1479–1487
to their injuries. Near house G (in Fig. 1a) one victim sustained 55% TBSA burn with respiratory burn and expired on 6th day.
3.5.
done to prevent communicable diseases. Debris from the site of accident on the road was cleared after police and legal formalities.
Resources 3.8.
Police chief and DMO reached on the spot within half an hour. Well placed disaster management plan was activated. Under direct supervision of Kannur District Collector, District Medical Officer, and Deputy District Education Media Officer-DMO Health a local team was formed that mainly people from a nearby village Edakadu. The prominent members of the team were: President, Edakadu Gram Panchayat, Medical District Education and Media Officer (DMO-Health) from Kannur, Health Inspectors from Edakadu Primary Health Centre. One hundred and fifty personnel from various fields (administration, Police, Fire force, residents from adjacent area) who were familiar to the place and people were required to manage the disaster. Initially victims were shifted to various hospitals by local people in their private or hired vehicle. Within 30 min 10 ambulances (including four High tech ambulances with facilities to transport critically ill patient hired from neighbouring district) were arranged to transport the victims. Initially victims were shifted to nearby hospitals where triage was done after first aid and then shifted to tertiary care hospitals in adjacent district and state.
3.6.
Difficulties faced
Problems in medical management of victims were experienced due to lack of high tech trauma centre, burns unit and ambulances in the state. It was difficult to manage crowd of volunteers gathered at the site and was managed by restricting their number. Local people objected to making videos of sufferings/ damages and uploading to social networking site immediately after the accident. As the people involved in rescue work initially from local known persons, no incidence by unscrupulous people was reported.
3.7.
1483
Rescue and relief work
Since communication and transport were not disrupted, rescue and relief work was relatively easy. Blast sound triggered immediate post-disaster response. Road was temporarily blocked and traffic were diverted to other routes. Availability of District Collector, District Medical Officer, and Police chief along with other officials on the site within 30 min for supervising and directing relief work made decision making very quick. Since the president of adjacent Gram Panchayat, was familiar with the place, people and custom of that area he was appointed as chief co-ordinator. Permanent District Information Centre and Temporary information centre set up by police served as information centre. Relief activities, local issues and victims condition were updated daily through mass media. Keeping in mind the available resources in the hospitals, victims were evenly distributed to all hospitals and clinics. All dead animals were removed from the area for safe disposal as per local predefined procedure. Phenol and Bleaching Powder were sprinkled; Water chlorination was
Social work
Appropriate compensation was declared from government and Indian Oil Corporation to minimize post disaster psychological problems, social disharmony and as a reassurance measure for the grieved relatives and victims.
4.
Mitigation measures
Following measures were taken to minimize risk in future:
(a) Divider was removed and road modifications done where ever suggested by competent authorities. (b) Security enhanced and traffic rules and regulations were strictly reinforced especially for vehicles carrying inflammable materials. (c) Emphasis was made to enforce standardized communication protocol, and safety protocol. (d) Rules related to essential drugs and supplies of emergencies were reviewed and strictly enforced. (e) Revamping of own government run burns and trauma centre along with high tech ambulances with critical care facilities and life support systems. (f) Liaison with external non-government and government agencies (administrative, health sector, relief related) were reviewed. (g) Exhaustive training and hi tech equipment to fire, rescue and police personnel to be given to deal with such disasters. (h) Education: timely cutting of electricity, Stopping use of mobile to avoid burst of LPG into flames.
5.
Discussion
Scientific literature on large scale LPG disaster is scanty in English literature. Majority of such incidences have been recorded in daily newspapers. Similar LPG burst has been mentioned in daily national newspaper (The Hindu, September 3, 2012, http://www.thehindu.com/news/states/kerala/article3854053.ece). The report mentions that on December 31, 2009, when an LPG tanker rammed a car at the Punthentheruvu junction at 3.50 a.m. and burst into flames. Twelve of the 18 persons who suffered burns died, most of them natives of Karunagapally and nearby areas who rushed in for rescue operations. Those killed were from the economically weaker sections, many of them daily wage earners. The tanker was carrying 18 tonnes of LPG from Kochi to the IOC’s refilling plant at Paripally. A major disaster was averted in Kochi on September 10, 2012 after 16 tonnes of cooking gas was accidentally released into the open at the Udayamperoor bottling plant of the Indian Oil Corporation (IOC), Kochi, Kerala. The Hindu, September 11, 2012 (http://www.thehindu.com/news/cities/
1484
burns 39 (2013) 1479–1487
Kochi/article3884493.ece) and The Times of India, September 11, 2012 (http://articles.timesofindia.indiatimes.com/201209-11/kochi/33762205_1_tanker-drivers-gas-leak-lpg). The accident occurred around 4.30 p.m. when the driver of a tanker moved the vehicle which was connected to the plant in a bay for transfer of gas. The valve broke, leaving no scope for plugging the leak. The entire quantity of gas in the tanker was released. The measures taken to avert a major fire accident due to LPG leak were: Stopping the movement of traffic for nearly an hour, cutting of electricity supplies. Residents in the vicinity of the bottling plant were alerted about the gas leak immediately and advised to leave their houses for safer places. In the tragedy of San Juanico (20 km North of Mexico City), on November 19, 1984 the disaster started due to LPG leakage, probably a pipe leakage or rupture due to excess pressure the storage tanks contained 11,000 m3 of a mixture of propane and butane (Pemex State Oil Company). The vapour cloud was ignited around 5:40 a.m. and was followed by an extensive fire at the plant area. The first explosion was registered on the seismograph at the University of Mexico at 05 h 44 min 52 s and was followed by a dozen explosions within the next hour, some of them of BLEVE type (Boiling Liquid Expanding Vapour Explosion) due to rupture of one or more storage tanks. Two of the explosions had an intensity of 0.5 on the Richter scale. Unburned and burning gas entered the houses south of the plant area and set fire to everything. Blast waves from the explosions not only destroyed a number of houses but also shifted several cylindrical tanks from their supports and added more gas to the fire. The smaller spheres and some of the cylinders exploded and fragments and even whole cylinders weighing around 30 tonnes were scattered over distances ranging from a few to up to 1200 m [2]. In the disaster reported in the present study the accident occurred when an Indian Oil Corporation LPG tanker rammed a road divider in an attempt to overtake a vehicle. The tanker containing LPG had three chambers (steel wall of 1 cm thickness) with total 16 tonnes (162.57 quintal) LPG in it. The LPG vapour was burst into flames within 20 min. The blast sound was heard up to 25 km distance and was soon followed by two more blasts at 3 min interval. After third blast approx one third part of container was thrown like missile 60–70 feet above the ground over coconut trees and later found about 500 m away from the site of accident in a field. LPG was seen by eyewitnesses flowing along the ground surface as LPG is heavier than atmospheric gases. The direction and speed of flow was determined by the direction of the wind and force of the burst. The coconut trees in the area were burnt up to top near the centre of the involved area, but the height of heat damaged portion of the trees gradually reduced at the periphery (Fig. 2 from Position A to more peripheral position B) indicating upper limit of heat making a shape of a dome. Since leaked LPG burns quickly and light weight heated gas expands over a wider area and moves upwards. Hence rapidly expanding burning LPG takes curved path from centre to periphery with convexity towards ground. Upward moving expanding hot gas contacts ground for shorter duration at the periphery (i.e. moves up towards periphery making a curve with convexity towards ground). This may be the possible explanation for relatively green vegetation at the periphery
(Fig. 2 at position B). Maximum death were recorded when victim positioned near seven feet high wall or near high house acting as block to reflect the blast wave and flow of hot air along with flame (Fig. 1a). Vegetation at the ground level at the periphery of affected zone showed little effect of heat suggesting that the lighter hot air and flame takes upward curve when force of blast wave becomes weak. Liquefied petroleum gas (LPG), which is used as a type of fuel, is stored as a liquid under high pressure in tanks. Immediate and sudden explosion of these tanks can release a large amount of gas and energy into the environment and can result in serious burns. Bozkurt et al. [3] have reported 18 patients injured due to LPG burns in five car LPG tank explosions incidents. The reported inhalation injury in 11 cases with varying degrees of severity, and 7 patients subsequently required mechanical ventilation. The explosions resulted from weakening of the tank wall (n = 2), crash impact (n = 2), and gas leakage from the tank (n = 1). In the LPG tanker explosion reported in the present study only one case of the six admitted in our burns centre had inhalation injury and required immediate ventilator support. The open space and distance of the victims could be the possible explanation for this difference in incidence of the inhalation injury.
5.1.
The pattern of burn (clinical forensic)
Based on the study of the victims in the field and admitted patient, the author found that the following points may be helpful to clinical forensic experts to further develop the clinical forensic in LPG burst burn (Fig. 4):
Fig. 4 – Pattern/type of burn at various positions in an open disaster area – Position A: extensive deep burn; sole burn if steps on the hot ground; Position B: extensive burn on the side of the body facing blast side and relatively deeper burn on lower positioned part of the body as LPG flows at the ground level; Position C: extensive but relatively superficial burn with deeper burn on the lower positioned part of the body facing the side of blast; Position D: less extensive mainly superficial burn on the part of body facing the side of blast; hot gases being relatively lighter starts moving upwards once the movement due to blast force become weak at some distance from the blast site and spares the relatively lower placed part of the body. Note: Sole burn may occur if patient steps on hot objects (thrown like missiles). As the flame disappears rapidly like flash, body parts covered by cotton cloth are usually spared. Respiratory burn may occur if victim’s position is near blast site or LPG is inhaled before catching fire.
burns 39 (2013) 1479–1487
1485
Fig. 5 – Patient (Case 3 in Table 1) at position A of Fig. 4. Her back was involved extensively.
(A) Victims near the site of accident in open area (if not protected by cloth or wall) develop extensive whole body burn. Burn is relatively deeper over lower positioned parts of the body (i.e. legs. hands, forearm) (Fig. 5). Pattern of depends on the type and material of cloth. Pattern is also determined by body posture and attempt to escape from the site. (B) Victims at the periphery of the fire zone (a) Mainly one side of burn indicates patient was exposed to fire for the short duration and probably he was positioned more towards the periphery. Victim’s position and posture may be determined by examining pattern of burn as more burnt surface of the body faces towards blast side (e.g. back involvement is more if victim is running away from the blast site, rt side body is involved more if victim tries to take right turn in an attempt to escape). (b) Foot burn is superficial/there will be no foot burn if the victim is positioned at the periphery in open area (Fig. 6). (c) At the extreme periphery (in open area) as the hot gases and fumes moves very fast towards upper strata, face, neck and exposed part of the body sustain superficial burn. Higher the position of the body part, more superficial burn it sustains (Fig. 7). (d) If victim is near a high wall even at the periphery of the affected zone, he/she may get extensive burn as expanding fires and hot gases gets reflected from the
Fig. 7 – Pattern of burn in a victim (Case 4 in Table 1) between position C & D of Fig. 4. Patient also had feet burns. Wounds were managed under LAD.
wall and produce turbulence in flow, in turn turbulence may induce movement of gases towards ground. The duration of exposure will also be more due to turbulence as well as due to failure of gases to spread to wider area (i.e. accumulation of more gas) due to blockade by the wall (Fig. 8). (C) Sole bun indicates walking on the hot ground near the site of accident or walking on the hot missile like objects due to blast. One of the victims the author examined at Periyaram Medical College had sole burn. History revealed that at the time of fire he was in his shop within 5 m from the site of the blast. (D) The significant difference in the level of burn of both lower extremities usually indicates the movement of fire and heat with blast wave (curved path with convexity towards ground) (Fig. 9). Usually the hand (upper extremity) will be involved to varying extent on the side the lower extremity is involved up to or above knee level. (E) Those who inhale gases or very close to blast are likely to develop inhalation injury (case 3 Table 1).
5.2.
Prevention of burn
Various studies have been done to find the model, mechanism and impact of boiling liquid expanding vapour explosion (BLEVE) to formulate the preventive strategies associated with BLEVEs [4–12]. Author would like to suggest following simple measures to avoid or minimize the extent of burn after
Fig. 6 – Patten of burn in a victim (Case 6 in Table 1) at extreme periphery (position D in Fig. 4) taking left turn to run away. Wounds were managed under LAD.
Fig. 8 – Pattern of burn in victim (Case 1 in Table 1) positioned close to a wall peripheral area. Wounds were managed under LAD.
1486
burns 39 (2013) 1479–1487
LAD design has notable advantages like wound isolation that reduces chance of wound colonization and safe disposal of infected materials (important factor to reduce hospitalacquired infections), while avoiding some major disadvantages such as opacity of dressing materials, inaccessible offensive smelling wound environment, and relatively high treatment costs. In LAD a definite intermittent negative pressure regimen is followed. The intermittent negative pressure (cycle of 30 min suction and 3½ h rest) is effective. Overall, the LAD is a safe and effective alternative to conventional dressing methods. LAD is an excellent research tool for wound healing as frequent/ continuous record of wound healing is possible without disturbing the wound healing process. LAD is an effective dressing for limb salvage in cases of acute and chronic complex wounds. Leech effect prevents wound related systematic response syndrome and sepsis.
Conflicts of interest There is no conflict of interest that could inappropriately influence (bias) my work.
Acknowledgements
Fig. 9 – Pattern of burn in a victim at position C of Fig. 4.
accident apart from measure taken at technology level and administrative level:
(a) Position behind the wall opposite side of accident will protect the body from exposure to heat. (b) One should move in open space away from the accident site to avoid getting trapped in turbulence due to reflected blast waves by obstructions (like wall) causing more extensive circumferential burn. (c) Covering the body with ordinary cotton sheet (and wearing cotton socks with pants tucked in it) while escaping may minimize the extent of burn. (d) One should leave the houses immediately for safer locations.
5.3. Use of limited access dressing (LAD) [13] during disaster Limited access dressing was found useful during disaster for managing large wounds. Limited access dressing (LAD) is a moist wound dressing with negative pressure. This is economical dressing that does not require frequent changing and saves resources and time. It provides limited access to the wound through two small ports for both dressers and pathogens. The
The author acknowledges the input by all eye witness (especially Mr. Bhaskaran P.K.) and support by Rajan Kelkar IAS (District Collector), Dr. R. Ramesh (District Medical Officer), Mr. K.N. Ajay (Medical District Education and Media Officer, DMO-Health, Kannur), Mr. C.H. Rajeevan (Health Inspector, Edakadu Primary Health Centre), Mr. Anish Babu (Health Inspector, Edakadu Primary Health Centre), Mr. K.V. Raveendran (President, Edakadu Gram Panchayat – Edakadu is the name of adjacent village), and administration of Pariyaram Medical College without which the study would not have been possible. The author also acknowledges the inputs by victims (especially Mr. Bhaskaran P.K., Mr. Kunhi Krishnan, Mr. Vineet Krishnan, and Mr. Kunhi Kanan) and relatives of the victims (especially Mr. A.K. Surendran) for extending all the help and providing photographs collected by them.
references
[1] Gupta A, Deepika S, Taly AB, Srivastava A, Surender V, Thyloth M. Quality of life and psychological problems in patients undergoing neurological rehabilitation. Ann Indian Acad Neurol 2008;11:225–30. [2] Arturson G. The tragedy of San Juanico – the most severe LPG disaster in history. Burns Incl Therm Inj 1987; 13:87–102. [3] Bozkurt M, Kulahci Y, Zor F, Kapi E. Burn injuries related to liquefied petroleum gas-powered cars. J Burn Care Res 2008;29:897–901. [4] Abbasi T, Abbasi SA. The boiling liquid expanding vapour explosion (BLEVE): mechanism, consequence assessment, management. J Hazard Mater 2007;141:489–519. [5] Ogle RA, Megerle MV, Morrison 3rd DR, Carpenter AR. Explosion caused by flashing liquid in a process vessel. J Hazard Mater 2004;115:133–40.
burns 39 (2013) 1479–1487
[6] Stawczyk J. Experimental evaluation of LPG tank explosion hazards. J Hazard Mater 2003;96:189–200. [7] Gong YW, Lin WS, Gu AZ, Lu XS. A simplified model to predict the thermal response of PLG and its influence on BLEVE. J Hazard Mater 2004;108:21–6. [8] Stawczy KJ. Experimental evaluation of LPG tank explosion hazards. J Hazard Mater 2003;96:189–200. [9] Papazoglou IA, Aneziris ON. Uncertainty quantification in the health consequences of the boiling liquid expanding vapour explosion phenomenon. J Hazard Mater 1999;67(June):217–35. [10] Shebeko YN, Bolodian IA, Filippov VN, Navzenya VY, Kostyuhin AK, Tokarev PM, et al. A study of the behaviour
1487
of a protected vessel containing LPG during pool fire engulfment. J Hazard Mater 2000;77:43–56. [11] Paltrinieri N, Landucci G, Molag M, Bonvicini S, Spadoni G, Cozzani V. Risk reduction in road and rail LPG transportation by passive fire protection. J Hazard Mater 2009;167:332–44. [12] Paltrinieri N, Bonvicini S, Spadoni G, Cozzani V. Costbenefit analysis of passive fire protections in road LPG transportation. Risk Anal 2012;32:200–19. http://dx.doi.org/ 10.1111/j.1539-6924.2011.01654.x [discussion 220–3]. [13] Kumar P. Exploiting potency of negative pressure in wound dressing using limited access dressing and suction-assisted dressing. Indian J Plast Surg 2012;45:302–15.