1111 2 A volcanic issue – lessons 3 4 learned in Goma 5 6 7 Aranka Anema and Jean-François Fesselet 8 9 10 1 When Mount Nyiragongo erupted in 2002 it sent lava 2 through the town of Goma and into Lake Kivu. Relief 3 operations focused on how to treat biological pollution 4 in water from Lake Kivu and to manage the water’s 5 chemical contamination from lava flows. Either form of 6 contamination would have been less of a problem if 7 lessons learned from earlier eruptions had been more 8 readily available. 9 20111 ount Nyiragongo in the water sources and unhygienic living that lava flow into the lake would 1 Democratic Republic of conditions. destabilize the gases and cause either a 2 M Congo is one of Africa’s fatal explosion resulting from methane 3 most active volcanoes. On 17 January Chemical and biological or the spread of lethal low-lying carbon 4 2002 Nyiragongo erupted, sending lava contamination dioxide, as happened by Lake Nyos 5 across the town of Goma 10km away in Cameroon in 1986. In addition to 6 into Lake Kivu. Goma has a population Lake Kivu is the only source of water disquiet over the destabilization of 7 of approximately 400 000. Years of for drinking, cooking and hygiene in carbon dioxide and methane in the 8 conflict, displacement and poverty have Goma and the surrounding areas. lake, specialists were concerned about 9 rendered Goma’s population particu- During the crisis, humanitarian agen- two other key water supply issues: the 30 larly vulnerable to disease and mal- cies became concerned that the water lake’s fluoride pollution and the threat 1 nutrition. Chief causes of mortality would not be safe for consumption of a cholera outbreak. Water quality 2 include malaria, measles, tuberculosis owing to chemical contamination. assessments and appropriate disinfec- 3 and diarrhoeal diseases. Epidemics such Following lava flows into the lake, tion techniques were considered crucial 4 as cholera, meningitis, dysentery and vulcanologists and water and sanitation to avoid wide-scale loss of life. 5 bubonic plague are frequent. Water and specialists noticed that the temperature Many factors must be taken into con- 6 sanitation infrastructure and health ser- of the lake had risen in localized areas sideration during assessments of water 7 vices in Goma are limited, and largely and was releasing large amounts of gas. quality, such as source protection, treat- 8 supported by external aid agencies. Dead fish floated to the surface, and ment efficiency and reliability, and 9 Because Goma’s health situation is flames reportedly glowed above the protection of the distribution network. 40111 already precarious, volcanic eruptions water. Vulcanologists studying Mount Water quality can be affected by 1 can easily prompt medical crises. In Nyiragongo have long known that Lake organic faecal pollution (e.g. bacteria, 2 1977 lava flow from Mount Nyiragongo Kivu holds large amounts of carbon viruses, protozoan pathogens, helminth 3 killed more than 2000 residents. In 1994 dioxide and methane. Scientists feared parasites, etc.), suspended matter 4 conflict and displacement triggered 5 cholera and dysentery epidemics, result- 6 ing in 50 000 deaths. 1 The eruption in 7 January 2002 led to the displacement of 8 over 300 000 to Rwanda and neighbour- 9 ing areas. Lava flows destroyed an esti- 50 mated 13 per cent of the town, including 1 water supply and electricity systems, 2 local health centres, hospitals, local resi- 3 dences and foreign aid offices. Goma’s 4 population suffered dehydration, res- 5 piratory and ocular infections, burns, 6 ionizing radiation and mental health 7 disorders as a consequence of the 8 eruption and subsequent displacement. 9 Risks of morbidity and mortality were 60111 heightened by the uncertain quality of Houses were submerged by the lava ows
Vol. 21 No. 4 April 2002 9 household water security
1111 (e.g. mineral solids, algae, protozoa, monitoring. This is particularly the case 2 bilharzia cercaria, etc.), acid and alka- for bucket chlorination. 3 line substances, toxic metals, pesticides Standard protocol for chlorine treat- 4 and nitrates, and concentrations of salts. ment calls for the preparation of a 1% 5 Of these, faecal pollution is generally chlorine stock solution, which is then 6 considered the most severe since it can mixed with raw water. An FRC of 7 lead to rapid and widespread outbreaks 0.2–0.5 mg/l of chlorine is considered 8 of communicable infectious diseases necessary in emergency settings in order 9 and mortality. to prevent possible recontamination 10 (e.g. during handling). Water pH levels 1 Cholera control and are known to influence the effectiveness 2 chlorination of chlorine disinfection, and chlorine 3 dosages must be modified according to 4 Drainage systems in Goma were badly pH levels; water with a high pH requires 5 damaged by the lava flows. Heavy rain- a higher FRC (0.4–1 mg/l). Lake Kivu 6 falls caused water pooling, flooding and has always had a high pH (8.5). Moni- 7 risks of associated faecal-oral diseases. toring local chlorination procedures 8 Cholera (infectious Vibrio cholerae ) has revealed that treated water did not have 9 been endemic in Goma for years; it is any FRC, either due to incorrect dilution 20111 transmitted through the ingestion of fae- procedures, a lack of monitoring equip- 1 cally contaminated food or water, and is ment or insufficient training. Local staff 2 preventable through a safe water supply had not always been properly trained 3 and hygiene measures. about assessment and monitoring tech- Chlorination requires properly trained staff and regular monitoring 4 The International Committee of the niques, and the need for regular modifi- 5 Red Cross (ICRC) had provided a local cations of chlorine content. Chlorination 6 water and sanitation NGO, Ami-Kuvi, points around the lake were using a 0.1 drastic increase to 4.0 mg/l. It is sup- 7 with training, finances and equipment per cent stock solution, rendering the posed that volcanic lava and acidic 8 in the past. During the crisis, Médecins treatment totally ineffective. The combi- ash affect the pH content of water and 9 Sans Frontières (MSF) began to collab- nation of poor chlorination techniques cause fluoride contamination. However, 30 orate with Ami-Kuvi (under agreement and the fact that infectious Vibrio scientific knowledge about the relation- 1 with ICRC) to monitor, train and cholerae proliferates in waters with high ship between lava, acidic ash and fresh 2 expand the NGO’s activities and to pH greatly heightened the risk of water toxicity are inconclusive, and the 3 ensure a safe water supply to Goma’s cholera outbreak. extent to which chemical modification 4 population. This involved the establish- In addition to concerns about cholera of water high in fluoride could be 5 ment of new chlorination points next prevention, WATSAN experts were also harmful to populations in the short 6 to the lake, and the rehabilitation preoccupied with the chemical (i.e. fluo- term is unclear. Water and sanitation 7 and monitoring of long-standing ones. ride) contamination of Lake Kivu. specialists working in Goma discussed 8 Trained staff performed bucket chlorin- several options: forbid everyone to 9 ation (i.e. on-site injection of chlorine Fluoride contamination drink water from Lake Kivu (unrealis- 40111 proportional to the capacity of individ- tic, given it is Goma’s only water 1 ual water containers). The World Health Organisation (WHO) source); allow everyone to drink lake 2 Chlorine is considered appropriate stipulates 1.5 mg/l as the safe level of water in spite of high fluoride content; 3 for the disinfection of water in fluoride content in drinking water. or treat the water with appropriate 4 emergency situations since it is power- Concentrations above 2 mg/l have been defluorination techniques. 5 ful enough to kill all viral and bacterial associated with the mottling of tooth Over a few days, water samples 6 pathogens. In order for disinfection to enamel (dental fluorosis) and concen- were taken at regular intervals and sent 7 be successful, however, it is imperative trations greater than 4 mg/l with joint to Rwanda for analysis. In the 8 that sufficient amounts of chlorine be problems and skeletal deformities. meantime, WATSAN specialists from 9 used. Chlorine is absorbed by oxidiz- Populations in hot climates are consid- humanitarian agencies held daily meet- 50 able substances present in water, such ered particularly at risk, since fluoride ings to exchange assessment findings 1 as organic matter, minerals, pathogens, concentrations in water are increased and discuss possible options. The most 2 etc. The effectiveness of chlorination is by evaporation. Furthermore, local diets frequently employed fluoride removal 3 dependent on the presence of the Free may already be rich in fluoride or nutri- techniques include ion adsorption 4 Residual Chlorine (FRC), and water tionally deficient. 2 (using activated alumnia or charred 5 quality must be assessed daily, or even Fluoride concentrations in Lake bone meal) and coagulation (using alu- 6 hourly, in order to adapt chlorine Kivu have always been slightly higher minium sulphate). Full-scale activated 7 dosages appropriately. Chlorination than the WHO recommendation, at alumnia facilities and household deflu- 8 requires good logistics (i.e. reliable 1.6 mg/l. A water sample taken by oridators using charred bone meal have 9 chlorine product, appropriate storage, Goma’s water control board a few been shown to decrease fluoride levels 60111 etc.), properly trained staff and regular days after the eruption revealed a from 5–8 mg/l to less than 1 mg/l. 3
10 Vol. 21 No. 4 April 2002 household water security
1111 employ new staff. Institutional memory 2 is often either held by individuals, or 3 contained in archives that are not acces- 4 sible to the public, and agencies are 5 forced to ‘re-invent the wheel’ every 6 time they encounter an unusual 7 emergency situation, such as the chemi- 8 cal destabilization of lake waters due to 9 lava flow. 10 Confusion over fluoride pollution 1 could have been avoided if WATSAN 2 experts had had access to the lessons 3 learned from earlier volcanic eruptions. 4 UNHCR had devised a contingency 5 plan about chemical water modification 6 and appropriate management techniques 7 after Mount Nyiragongo’s previous 8 eruption, but this document was never 9 readily available to all organizations. 20111 Inter-agency collaboration and commu- 1 nication during the humanitarian crisis 2 was excellent. However, humanitarian 3 organizations also need to take a shared 4 responsibility for the assimilation, 5 Water from Lake Kivu is used for bathing, washing and cooking recording and dissemination of relevant 6 operational information, so that 7 decision-making and programme imple- 8 Meanwhile, coagulation using restoration to health and restore dignity. mentation in future emergency contexts 9 aluminium sulphate can reduce fluoride Implicit in the humanitarian approach is is efficient. 30 content by 10–60 per cent. 4 These tech- a ‘quick in and quick out’ response, 1 nical options were never discussed in which means there is a tendency to References 2 detail, however, since they would have overlook the need for transferring 3 been logistically unfeasible in the con- knowledge and technical expertise to 1Goma Epidemiology Group (1995) ‘Public health impact of Rwandan 4 text of Goma. Furthermore it was relevant local NGOs. refugee crisis: What happened in Goma, 5 assumed that the impact of water high The incomplete chlorination assess- 6 Zaire, in July 1994’, The Lancet, 345: in fluoride content on the population ment, dilution and monitoring 338–43. 7 would not be immediate. techniques witnessed in Goma could 2Feachem, R.G. and Cairncross, S. 8 Before humanitarian agencies had have been prevented if local NGOs had (1993) Environmental health engineer- 9 the time to discuss an appropriate solu- received adequate training in, and mon- ing in the tropics: an introductory text, 40111 tion, new water samples showed a itoring of, water disinfection. Special 2nd edn, John Wiley, London. 1 decrease in fluoride concentration to attention needs to be paid to explaining: 3WHO (1993) ‘Protection and improve- 2 2.0 mg/l. The agencies decided to leave ment of water quality’, in Guidelines for 3 the chemical content of the water why chlorination techniques are used drinking-water quality, 2nd edn, Vol.1 – 4 untreated. in emergency contexts Recommendations , WHO, Geneva. 5 what factors influence disinfection 4Twort, A.C., Ratnayaka, D.D. and Brant, M.J. (2000) Water supply , 5th 6 protocol (e.g. pH levels, turbidity, Lessons learned edn, Arnold Publishing, London; and 7 presence of iron and organic matter, WHO (1993) Guidelines for cholera 8 Confusion about the consequences of etc.) control, WHO, Geneva. 9 fluoride contamination and incomplete which types of biological pollution 50 chlorination procedures brought home chlorine is capable of eliminating 1 several lessons: (e.g. bacterial and viral) and not Further reading 2 International protocols and appro- (e.g. protozoa, helminths, etc) Mann, H.T. and Williamson, D. (1982) 3 priate training for bucket chlorination. how to disinfect water properly (e.g. Water treatment and sanitation, ITDG 4 Unlike sustainable development organi- pH assessment, contact time, calcu- Publishing, London. 5 zations, humanitarian organizations are lation of FRC, etc.). 6 not focused on long-term capacity About the author 7 building and participatory programme Humanitarian agencies need to Aranka Anema is a medical anthropologist, currently 8 implementation. The goals of MSF, for strengthen their institutional memory. working with MSF-Holland as Medical Editor. Jean- 9 example, are to prevent widespread Relief agencies constantly encounter François Fesselet is the Co-ordinator of the Water and 60111 mortality, alleviate suffering, prompt a new emergency situations and have to Sanitation Unit at MSF, Amsterdam.
Vol. 21 No. 4 April 2002 1 1