The Sanitation Safety Planning methodology is implemented within a cooperation project in Iringa, Tanzania. The study presents the methodology and its adaptation and use for the given context, in order to assess risks and to support stakeholders in improving the current sanitation system and validate the design of an improved one. First results of the application of the methodology, obtained in one of the four peri-urban wards of Iringa, demonstrated its efficacy and utility in prioritising risks and identifying cost-effective control measures. Risks were assessed by the use of a semi-quantitative approach, and a simplified risk assessment matrix was developed for the case study. A sensitivity analysis was carried out in order to evaluate criteria for prioritising control measures to be selected for the development of an achievable improvement plan.
The importance of wastewater and fecal sludge management and reuse is increasing as well as the evidence of the relation between an appropriate sanitation and the protection of public health (Mara et al. 2010). Despite the evidence of the problem and international efforts, progresses to improve the situation are slow particularly in developing countries, due to many factors including lack of expertise, limited budget and poor political will (Cairncross et al. 2010). There is a need for providing safe affordable sanitation systems, increasing awareness and coordinating efforts of concerned stakeholders. The World Health Organization (WHO) developed the Sanitation Safety Planning (SSP) methodology, a risk-based approach aimed to give support to sanitation operators to assess and minimise health risks related to sanitation systems (WHO 2015).
The SSP methodology is under implementation within the cooperation project ‘Integrated Environmental Sanitation Concepts for Poor, Underserved and Peri-Urban Areas of Iringa Municipality’, founded by the European Union and implemented by the Italian non-governmental organisation (NGO) Fondazione ACRA, with CeTAmb LAB and BOKU as technical partners. The project aims at improving health and hygiene conditions of poor communities living in four peri-urban wards of Iringa, Tanzania. According to results of the baseline survey conducted in the targeted area, 45% of household latrines are considered unimproved (UNICEF & WHO 2015), meanwhile most of the wells show evidence of fecal contamination. Mortality of children under five years caused by diarrheal diseases is 16.7% in peri-urban wards. Almost 44% of households rely on informal emptying and fecal sludge is often reused in agriculture without treatment. In this scenario, the proper management of the sanitation system represents a priority for safeguarding the health of communities and workers of the sector (Bartram & Cairncross 2010; Mills & Cumming 2016).
The paper focuses on the testing and adaptation of the SSP methodology for the Iringa case study. The implementation of the SSP methodology was foreseen to support technicians and local authorities by risks identification and prioritisation in the current sanitation system and by the conception and management of an improved sanitation system. An improved sanitation system was developed by the support of the SSP methodology assessing the appropriateness of proposed solutions in lowering health risk and in developing an incremental improvement plan for the target area.
Since few case studies are published in the international literature, and due to the fact that sanitation systems are widely complex and involve many stakeholders, the need for further validation of the SSP approach emerged. The literature shows a wide number of studies on health risks related to fecal sludge management and reuse, as well as measures for their control (Blumenthal & Peasey 2002; WHO 2006; Stenström et al. 2011). Nevertheless, few publications are focused on SSP implementation (Jackson & Vuong 2014; PAHO 2015; WHO 2015; ADB 2016; Nicolics & Langergraber 2016), even if some case studies were presented in international conferences (Water Safety Conference, Uganda, 2012; 12th Sanitation Community of Practice SanCoP, UK, 2013; IWA Water and Development Congress, Jordan 2015). Besides presenting a case study for the implementation of the SSP methodology, an additional objective of the paper is to present a simplified methodology for risk assessment in data poor environments. To prove its applicability, the new risk assessment methodology is compared with the original matrix proposed in the WHO manual.
Use of the SSP methodology
This study is based on the methodology described in the WHO SSP Manual (WHO 2015) and was carried out by authors at research level. In Iringa, the assembling of an SSP team is still under progress and has not been accomplished yet, owing to the research stage in which the SSP method is tested at this point. Data were derived from literature, official documents of local authorities and institutions and project reports. Field data were collected during the survey carried out in November 2014 by means of interviews and direct observations made with 718 households in peri-urban wards, interviews to stakeholders and field visits to the Wastewater Municipal Treatment Plant (WWTP) and to sites interested by pit emptying. The study was a baseline survey endorsed by the District Executive Director and the District Medical Officer. Informed consent was obtained from participants. Analytical data were collected through water monitoring campaigns conducted at the WWTP and in wells located in the project area.
The paper focuses on the implementation of the SSP methodology in the peri-urban ward of Kitwiru, which is located in the south of Iringa Municipality and has about 18,200 inhabitants (NBS & IMC 2013). First, the SSP was applied to the current sanitation system, hereafter named S0. Hazards and their related hazardous events were identified for each step of the sanitation chain, as well as exposure groups and transmission routes. Because of the scale and the scarce availability of site-specific data on pathogens, they were considered as a unique category defined as ‘all microbial pathogens’, even if distinct types could result in diseases with different levels of severity, and some of them are more resistant than others to treatments (Roma & Pugh 2012).
Risks related to each step of the sanitation chain in S0 were assessed using a semi-quantitative approach. Existing control measures were not considered during the scoring of risks; they were evaluated in a second stage and additional controls were identified for events characterised by high risks.
Improved sanitation system
The CLUES approach was applied within the project according to guidelines by Lüthi et al. (2011), in order to design a new improved sanitation system (hereafter named S1) for peri-urban wards. The approach allowed the identification of preferred and implementable technologies and systems through a participatory process involving beneficiaries and other stakeholders.
Additional control measures as emerged from the SSP approach applied to S0, if not already present, were included in the design of the improved system S1. The SSP methodology was applied to S1. The risk assessment for S1 was evaluated using the simplified matrix. Control measures were hypothesised to be in place and operational.
Hazards for S0 were first assessed by means of the matrix of Likelihood (Li) and Severity (S) as suggested in Module 3 of the SSP manual (WHO 2015), where Li assumes scores from 1 (very unlikely) to 5 (almost certain), while S from 1 (insignificant) to 16 (catastrophic).
Due to the complexity of sanitation systems, a simplified matrix has been developed to streamline the risk evaluation procedure, and to facilitate the work of the SSP team. Table 1 illustrates the simplified matrix used in this work, with severity and likelihood definitions elaborated for the current case study. Li assumes scores from 1 (unlikely) to 3 (almost certain), whilst S from 1 (minor) to 6 (major). Risks are categorised as low (L), medium (M), high (H) or very high (VH). A high level of risk (H) is associated to hazardous events with a total score (Li • S) greater than 5.
The same risks for S0 were re-assessed by the use of the simplified matrix to compare its sensitivity with the one suggested by WHO.
Prioritisation of control measures
Particular attention was given to the prioritisation of interventions and the assessment of their efficacy, which have great importance in a low-income context. A sensitivity analysis of control options was carried out based on Module 4 of the WHO SSP manual (WHO 2015) for prioritising interventions in the improvement plan. Priority scores were calculated on the base of four selected criteria, defined as follows:
Potential (P): human capacity of improving existing controls;
Technical Effectiveness (TE): efficacy of controls in reducing risks;
Acceptability (A): reliability of controls in relation with local cultural and behavioural habits;
Costs (C): budget necessary for installation and implementation of controls.
RESULTS AND DISCUSSION
Implementation of the SSP methodology
The improved sanitation system S1 foresees the use of ventilated improved pits, twin pits pour flush toilets, septic tanks, or fossa alterna. The emptying and transport would be operated manually or mechanically, whilst sludge would be disposed of at a decentralised co-composting plant to produce compost for agriculture, or at the WWTP. The stabilised humic material obtained in the fossa alterna could be used as a soil amendment (Tilley et al. 2014). Table 2 shows the risk assessment applied to the current situation S0 and hazardous events characterised by a high score related to the sanitation chain step mechanised emptying, selected as an example. Complete results are shown in Table S1 (Supplementary material, available with the online version of this paper). A total of 17 risks out of the 103 identified obtained a high or VH level score. High risks occurred in each step of the sanitation chain and concerned all exposure groups, principally vulnerable people, workers dealing with manual emptying, farmers reusing untreated fecal sludge and consumers of raw products.
The assessment of existing control measures has emerged as a challenging aspect of the SSP methodology. A sanitation system at the ward scale involves a broad number of stakeholders, in particular users and local community members. Some of the control measures imply personal behaviour and habits, such as washing hands, wearing shoes, the preparation of food, etc. It has been shown that the assessment of risks is easier for systems where the process can be better monitored (e.g. if a composting site is considered (Jackson & Vuong 2014)). On the ward level, the assessment of each step and control measure is much more difficult. Therefore, combined with the lack of site-specific and reliable data on hygienic behaviour, the evaluation of the existence of control measures and their effectiveness in reducing risks could be misrepresentative. As a result, risks were assessed without considering control measures in place. Existing controls for each hazard were listed, evaluated and discussed afterwards. On the basis of these considerations, additional actions were suggested for those risks obtaining high scores.
Improved sanitation system
A total of 38 control measures were identified and included in the design of S1. When the SSP methodology was applied to S1, hazards and hazardous events resulted in risks with a low or medium score, showing that adequate measures can effectively control the risks and verifying the efficacy of the designed system in addressing high level risks.
The improved system S1 is based on a multi-barrier approach focused on additional control measures suggested for risks reduction in S0:
strengthen awareness campaigns for improving hygiene behaviour;
specific trainings of emptiers, aimed at increasing awareness and improving practices in using appropriate PPE and tools;
improved latrine design and faecal sludge management technologies;
enforcement of law and regulations in the sanitation sector.
V (=verified) and X (=not verified) indicate if the risk level obtained using the two matrices remained the same or not. Complete results are included in Table S2 (Supplementary material, available with the online version of this paper).
Scores showed a good match when using the simplified matrix: of the 103 risks assessed for S0, the 83% of risk levels could be verified using the simplified matrix. Some risks, with low Li and moderate S, gained low instead of medium level of risk using the simplified matrix. For SSP objectives of this research, focused on identifying and controlling high risks, this change was not considered significant. Also, some risks having low Li and highest S were classified as high instead of medium. Unifying some Li and S values caused a loss in sensitiveness in the simplified matrix. In consideration of precautionary principle and the importance of evidence risks with highest severity (including loss of life) the sensitivity of the simplified matrix was considered adequate.
Prioritisation of control measures
Results of the sensitivity analysis on control options identified for high level risks showed that controls are sensitive to the modification of weight assigned to criteria (Table 4). Sixty-three percent of control measures change the priority according to the variation of weights. Cost is the criterion that influences the prioritisation of control measures more than others. For the case study, it was chosen to assign a higher weight to Potential and Costs criteria in order to encourage the adoption of less capital intensive solutions in the short term; the Acceptance criterion was given a low value in consideration of the importance of awareness increasing expressed by local community members and authorities, convinced of its efficacy. The results of the assessment with the weights proposed for the case study are shown in column w(S0) in Table 4.
Scenario w(1): all criteria have the same weight equal to 0.25.
Scenario w(P): P criterion has a double weight (wP = 0.4) respect to others (wTE = wA = =wC = 0.2).
Scenario w(TE):TE criterion has a double weight (wTE = 0.4) respect to others (wP = wA = wC = 0.2).
Scenario w(A): A criterion has a double weight (wA = 0.4) respect to others (wTE = wP = wC = 0.2).
Scenario w(C): C criterion has a double weight (wC = 0.4) respect to others (wTE = wA = wP = 0.2).
Scenario w(S0): Weights assigned to criteria for the case study are wP = 0.3, wTE = 0.2, wA = 0.1, wC = 0.4.
Differently from the case study of Parque Huàscar, Peru (PAHO 2015), the Costs criterion was added for calculating the priority of interventions, as well as weights being attributed to each parameter. This shows the importance in the given context of including the cost as a criterion for supporting the identification of cost-effective interventions.
The SSP methodology demonstrated effectiveness for identifying hazardous events with associated high risks and cost-effective interventions in Iringa. The methodology helped to study in depth the current sanitation system and to identify hazardous events with associated high risk. These results permitted to evidence additional control measures which were integrated in the proposed improved sanitation system, based on a multi-barrier approach. The application of the SSP methodology supported the development of an improvement plan in which interventions are planned and prioritised to maximise the health protection of communities and workers, promoting the recovery of resources. The same benefit was reported by Jackson & Vuong (2014) who applied the SSP methodology in Hanoi, Vietnam.
The developed simplified matrix for evaluating risks was sensitive enough with regard to project objectives. The limited set of values proposed for the simplified matrix facilitates the work of local SSP team members not experienced with risk assessment.
The sensitivity analysis showed that weighting of criteria for control measures evaluation has an influence on prioritisation of options. Thus, weights should only be assigned after discussion by the SSP team.
The results of testing the SSP methodology to one ward represent the starting point for the use of the methodology and to extend SSP activities to other wards in the municipality, in order to support project activities in improving the Iringa sanitation system. Moreover, Tanzanian institutions and utilities showed a strong interest in the SSP approach. Therefore, there is potential for scaling up SSP application in Tanzania.
Additionally, the simplified matrix should reduce efforts for implementation of the SSP methodology in similar situations with, for example, lack of data, severe economic constraints, and scarce experience and familiarity of practitioners with risk assessment.
The authors are grateful to the whole ACRA Foundation staff who strongly supported the implementation of the research. Mr Maswaga supported data collection and elaboration, while Prof. Langergraber collaborated in the study and implementation of the SSP methodology and in the review of the paper. Eng. Rondi e Prof. Sorlini supervised the whole work in its progress. All offered valuable suggestions for this paper that are gratefully acknowledged. The views expressed do not necessarily reflect those of the funding organisations.