The intervention
TIMELINE
WHAT WAS DEPLOYED
Between 2003 and 2019 Indonesia built 21,832 neighborhood-scale decentralized wastewater systems under the SANIMAS approach, at a cumulative investment of about US$941.6 million (nominal, 2002 to 2019), an average of US$32,505 per site [3]. Systems are passive anaerobic treatment units (settler plus anaerobic baffled reactor, later with anaerobic filter) serving on average 54 household connections, delivered either as communal toilet and washing blocks (MCK) or simplified sewer systems, then handed to volunteer community groups for operation [1][3]. The treatment technology largely worked: 92% of 99 tested early systems met the then-applicable BOD standard of 100 mg/L [1]. Use did not follow design. A 298-system survey in 2011 found median usage of 100 people per site against a design basis of 250 to 500 [1], and about 10% of facilities were dysfunctional and not in use [1]. National monitoring reported average utilization below half of design capacity, one third of systems without an active management group, and 80% never desludged [2]. The 2021 independent evaluation of the IsDB-financed cohort (236 systems surveyed, with onsite inspection and laboratory effluent sampling at a subset) rated overall performance poor at 51% and of concern at 43% of sites; 37% of community groups had no income source and 20% had no operator [3]. Communal blocks fell out of use as households acquired private toilets [1][3]. The build program continued; individual systems were progressively abandoned by users after handover [2][3].
HOW IT WORKS
Physical: passive anaerobic treatment at neighborhood scale, typically a settler followed by an anaerobic baffled reactor (two thirds of systems), later designs adding an anaerobic filter; no moving parts and no energy input; effluent discharges to open drains, canals or rivers. Wastewater arrives either from a communal sanitation center (toilets, bathing, laundry; 77% of the 2003-2010 cohort) or via a simplified gravity sewer from household toilets (16%), or both (6%) [1]. Administrative: national government (APBN and DAK budget lines) or development bank loans (ADB, IsDB) grant construction funds directly to a neighborhood community group (KSM), which builds the system with NGO or consultant facilitation; on completion, operation transfers to a volunteer beneficiary group (KPP) expected to appoint an operator, collect user fees, and fund operation, maintenance and desludging indefinitely [1][2][3].
The measurements
OUTCOMES
| METRIC | VALUE | METHOD | MEASURED BY | INDEPENDENT | PERIOD | SRC |
|---|---|---|---|---|---|---|
| Sites rated poor overall performance (IsDB 2014-2019 cohort) | 51 % of surveyed sites | Performance-indicator survey: structured phone interviews with 236 community operating groups, onsite inspections by five field teams, field measurements and certified-laboratory effluent sampling at a subset | CWIS TA-Hub consultant team for Bappenas, IsDB and BMGF | ● | 2020 | [3] |
| Sites with one or more serious management problems (37% no O&M income, 20% no operator, 7% abandoned before completion, 3% major physical damage) | 48 % of surveyed sites | Same survey | CWIS TA-Hub consultant team | ● | 2020 | [3] |
| Median users per system versus design basis | 100 persons per site (design basis 250-500) · baseline 250-500 (design) (at commissioning) | Survey of 298 systems operational at least one year (BORDA-WSP 2011) | WSP-commissioned survey co-executed with implementing NGO BORDA | ○ | 2011 | [1] |
| Facilities completely dysfunctional and not in use (2003-2009 cohort) | 10 % of facilities (approximate) | Site surveys and visits across nearly 400 sites | WSP review team | ● | 2011-2012 | [1] |
| Effluent BOD compliance with then-applicable 100 mg/L standard (ABR systems) | 92 % of 99 tested sites | Effluent sampling at ABR sites (BORDA-WSP data), consistent with 2011 Ministry of Public Works tests at 22 sites | BORDA-WSP; Ministry of Public Works | ○ | 2011 | [1] |
| Measured COD effluent concentrations not complying with 2016 national discharge standard (100 mg/L) | 65 % of measurements | Onsite sampling with certified-laboratory analysis | CWIS TA-Hub field teams | ● | 2020 | [3] |
| Systems operating longer than 2 years never desludged (IsDB cohort) | 71 % of systems | Structured operator phone interviews with photo verification | CWIS TA-Hub | ● | 2020 | [3] |
| Average utilization relative to design capacity (national monitoring) | <50 % of design capacity | National monitoring data of community-scale systems as reported in peer-reviewed synthesis | AKSANSI/ISF-UTS monitoring cited by Willetts et al. | ● | circa 2016-2018 | [2] |
| Systems without an active management group | 33 % of systems (reported as one third) | National monitoring data | cited by Willetts et al. | ● | circa 2016-2018 | [2] |
| Community groups collecting any user fees | 60 % of groups | National monitoring data | cited by Willetts et al. | ● | circa 2016-2018 | [2] |
| Investigated communal blocks (MCK) with observed functional problems | 70 % of investigated MCKs | Operator phone interviews and onsite observation | CWIS TA-Hub | ● | 2020 | [3] |
| Sites rated poor where operating group is trained and funded and operator paid | 11 % of such sites (versus 51% overall) | Cross-tabulation within the same performance-indicator survey | CWIS TA-Hub | ● | 2020 | [3] |
● independently measured○ operator-reported· not stated in any source
COST BASIS
- CAPITAL
- USD 941,582,000 (2021)
- OPERATING
- not published
- PER UNIT
- 32505 USD per system location, average construction cost 2002-2019, as reported 2021 [3]
- FUNDING
- 3
- 3 · 100000000
- 3
- 3
- 3
Capital figure covers construction grants per location compiled from program records; inclusion of facilitation and program overhead costs is not fully specified in the source. Median capital cost per person served in the 2003-2009 cohort was US$294 for communal sanitation centers and US$228 for simplified sewer systems (2011 survey data); low usage inflated per-capita cost above design intent [1]. Household fees where collected were typically US$0.18-0.31 per month [2].
The assessment
PRECONDITIONS
- Dense low-income urban settlement where on-site sanitation is infeasible and no sewer service exists or is planned soon [1]
- Warm (tropical) climate suitable for passive anaerobic treatment [1]
- Gravity drainage to the treatment site and a legal discharge point [1]
- Community with sustained voluntary-service ethic and stable leadership; systems decayed where leadership changed [1]
- Local government post-construction support: monitoring, technical help, and a desludging service, since communities do not regard desludging as their responsibility [1][2]
- A trained and paid operator and a funded operating group; performance was markedly better where these existed [3]
- Secure land tenure and resolved asset ownership for the system [2][3]
- Formalized fee setting and collection with authority behind it [2]
- For communal blocks: siting only where households genuinely cannot build private toilets, otherwise usage collapses as piped water spreads [1][3]
FAILURE ANALYSIS
The dominant failure mechanism is institutional and social, not treatment technology. The design assumption that volunteer community groups would operate, finance and maintain wastewater infrastructure indefinitely after a construction grant is described by the WSP review as overstated: groups lose enthusiasm, desludging is neglected, and even committed users will not fund major repairs [1]. Documented mechanisms: (1) Volunteer management collapse: one third of systems nationally had no active management group [2]; in the 2021 evaluation 37% of beneficiary groups had no income source for O&M and 20% had no assigned operator [3]; over half of operators worked without cash payment, driving turnover and loss of training [1]. (2) Fee collapse: only 60% of groups collected any fees [2]; more than a quarter of communal centers had no regular income at all [1]. (3) Non-connection and disuse: households did not connect their toilets, some not realizing that was the intended purpose [2]; 75% of surveyed operating groups reported households declining connection, and 31% of interviewed communities reported disconnections out of dissatisfaction or house renovation [3]. (4) Obsolescence by substitution: communal toilet blocks lost users as piped water spread and households built private toilets, the leading documented reason MCKs fall out of use; 70% of investigated MCKs had functional problems and the evaluation judged 80% may not be sustainable [1][3]. (5) Absent post-construction support: local government monitoring, intervention and desludging were largely missing; 71% of systems older than two years had never been desludged [1][3]. The strongest evidence that the failure is social rather than technical: where operating groups were trained and funded and operators paid, sites rated poor fell from 51% to 11% [3].
Stated in the reviews themselves. WSP 2013 [1]: reconceive community management as co-management ("Sanitation with Communities"), with communities handling above-ground daily operation and local government responsible for below-ground sustainability, desludging, sludge disposal and major maintenance; make simplified sewer systems, not communal blocks, the default wherever conditions permit; develop systems only within a citywide sanitation plan; secure a pre-construction memorandum of understanding on roles and long-term costs at every site; size systems for at least 100 households. Willetts et al. 2020 [2]: four minimum local government responsibilities: monitor and keep records on all systems, provide post-construction technical and institutional support, formalize fee setting and collection, and fund major repairs and rehabilitation; revise national guidelines on responsibility allocation, tariffs, asset ownership and eligible budget lines. The 2021 independent evaluation [3]: 15 recommendations including resolving asset ownership with cities and communities, training and paying operators (associated with poor ratings falling from 51% to 11%), varying the 2016 effluent standards to balance cost and coverage, integrating SANIMAS with scheduled desludging services, and setting national-to-local key performance indicators.
TRANSFERABILITY ASSESSMENT
| AXIS | SCORE | RATIONALE |
|---|---|---|
| CAPITAL INTENSITY | ●○ | About US$32,505 per system is small individually, but rollout depended on national fiscal transfers (DAK) and development bank loans, a capital program rather than routine municipal spending [3]. |
| GOVERNANCE DEPENDENCY | ●○ | Delivery used existing community-grant mechanisms, but evaluations found asset ownership, community group legal status and budget-line authority unresolved, requiring regulatory change to fix [2][3]. |
| ENVIRONMENTAL SPECIFICITY | ●○ | Passive anaerobic treatment depends on warm ambient temperatures and dense settlements with gravity drainage; broadly applicable across tropical cities, constrained in cold climates and low-density areas [1]. |
| SKILLS AND MAINTENANCE | ○○ | The documented failure mechanism: volunteer groups could not sustain operation, financing and desludging without ongoing external support; performance was acceptable mainly where operators were trained and paid and government supported post-construction [1][2][3]. |
| FUNDING REPEATABILITY | ●○ | Capital was repeatedly funded through recurrent budget lines, but lifecycle funding failed: user-fee recovery collapsed and no budget line reliably covered desludging and major repairs [1][2][3]. |
| POLITICAL DURABILITY | ●● | The approach expanded from 6 pilot sites to 21,832 systems across successive national administrations, 2003-2019 [3]. |
TOTAL 6/12 · T3
POLITICAL TEST
Survived national political transitions repeatedly: piloted 2002-2004, adopted 2005-2010, expanded 2011-2019 across successive administrations and six funding programs, growing from 6 pilot sites to 21,832 systems with almost US$1 billion invested [3]. The failure documented here occurred at community scale after handover, through disuse and management collapse, not through withdrawal of national political support. This is why the record is classified failed_social rather than discontinued_political.
The file
SOURCES
- [1] post_implementation_review · Review of Community-Managed Decentralized Wastewater Treatment Systems in Indonesia (WSP Technical Paper) · Kathy Eales, Isabel Blackett, Reini Siregar, Evi Febriani · Water and Sanitation Program, World Bank (study funded by AusAID) · 2013-06 · en · PRIMARY · INDEPENDENT · bdd.pseau.org Full PDF fetched and read. Canonical record also at openknowledge.worldbank.org/handle/10986/17751 (returned 403 on access). The underlying 298-system survey (BORDA-WSP 2011) involved the implementing NGO BORDA, so survey-derived figures are only partially independent.
- [2] peer_reviewed · Sustaining Community-Scale Sanitation Services: Co-management by Local Government and Low-Income Communities in Indonesia · Juliet Willetts, Freya Mills, Mova Al'Afghani · Frontiers in Environmental Science, vol. 8, article 98 (doi:10.3389/fenvs.2020.00098) · 2020 · en · PRIMARY · INDEPENDENT · www.frontiersin.org Open access; fetched and read. Case studies in Kota Bogor and Kabupaten Bantaeng, 2016-2018, plus national monitoring figures.
- [3] post_implementation_review · Independent Evaluation of SANIMAS Model as an Approach for Providing Decentralised Sanitation · Paul Bulson, Susilawati Muslimah, Nicolas Reynaud, Isabel Blackett, Alexander Viwat Campbell, Inni Arsyini · City-Wide Inclusive Sanitation Technical Assistance Hub, for Bappenas and the Islamic Development Bank; funded by the Bill & Melinda Gates Foundation · 2021-02 · en · PRIMARY · INDEPENDENT · www.isdb.org Full 115-page PDF fetched and text-extracted; forewords dated Jakarta, February 2021, hosted by IsDB March 2021. Fieldwork August-late 2020: 236 remote structured operator interviews, onsite inspections by five field teams, certified-laboratory COD/BOD analysis, FGDs planned for 6-8 beneficiaries in 50 locations.
RECORD
- PUBLISHED
- 2026-08-08
- LAST VERIFIED
- REVIEW DUE
- 2027-02-07
- RUBRIC
- v1.0
- STATUS HISTORY
- · failed social · Record published.
- RIGHT OF REPLY
- Not yet exchanged. The failure analysis in this record rests entirely on evaluations commissioned and published by the program's own principals (World Bank WSP 2013; Bappenas, IsDB and BMGF 2021) and on peer-reviewed research. Right of reply will be offered to the Ministry of Public Works and Housing and Bappenas before this record is promoted or distributed, and any response received will be published here in full.
- LANGUAGE
- All sources are English-language publications. Indonesian program names (SANIMAS, KSM, KPP, MCK, DAK, APBN) are used as in the sources.
- CONFIDENCE
- Solid: program scale (21,832 systems to end-2019, 97% by MoPWH), cumulative investment (US$941.582 million nominal), average cost per location (US$32,505), IsDB financing (US$100 million approved 2013), and all utilization, management, desludging and effluent figures were read directly from the cited documents, not from abstracts or secondary summaries. Thin or judgment-based: (a) the single outcome_status label. Outcomes are heterogeneous: the 2013 review found over 80% of the 2003-2009 cohort functioning day to day, while later and larger evidence shows majority underutilization and management collapse; failed_social reflects the documented pattern of technically functional systems progressively unused after handover, and applies most strongly to communal toilet blocks; "degraded" would be a defensible alternative label. (b) population_served (6 million) is the 2021 evaluation's estimate derived from systems and average connections, not a user census. (c) System counts differ across sources: about 30,000 by end-2017 in [2] (broader definition including other community-scale programs) versus 21,832 SANIMAS-program systems to 2019 in [3]; the compiled program table from [3] is used. (d) Pilot start differs: 6 sites 2002-2004 [3] versus seven sites 2003-04 [1]; timeline uses year precision only. (e) No program-wide annual operating cost exists; operating_annual left null. (f) The 92% BOD compliance figure comes from a survey co-executed by the implementing NGO and is marked not independent. (g) The 2011 survey's median-usage baseline (design 250-500 persons) is a design intent, not a measured baseline. No figures were estimated or interpolated by the registry; all nulls reflect absence in sources.
CITE THIS RECORD
change4planet.org. C4P-WTR-0001: Community-managed decentralized wastewater treatment (SANIMAS), Indonesia: post-handover disuse at national scale. Grades T3/E2 (rubric v1.0). Last verified 2026-08-07. https://change4planet.org/records/c4p-wtr-0001-sanimas-indonesia