Groundwater in Bangladesh: Depletion, Arsenic, and Sustainability

Bangladesh extracts about 30.2 BCM of groundwater a year against renewable recharge near 21.2 BCM, an overdraft of roughly 9 BCM, about 42% above recharge (Master Planning Organization 1997; FAO 2013, via Planning Commission). The gap against sustainable yield is wider still, because a share of recharge must stay in the ground to sustain dry-season baseflow and ecosystems. The depletion is regional and self-reinforcing: the Barind Tract falls about 0.30 m a year on average and 1 to 2 m in hotspots, while Dhaka draws most of its supply from a single stressed aquifer. The highest-leverage action is to put a price on extraction before the SREDA and IDCOL solar irrigation rollout drives the marginal cost of pumping toward zero. Permitting under the Water Act 2013, a phased move toward the WHO arsenic standard, and metered solar pumping are the three moves that change the trajectory.

Executive summary

• Groundwater overdraft is about 9 BCM per year, roughly 42% above renewable recharge (about 30.2 BCM extracted vs roughly 21.2 BCM recharged; MPO 1997; FAO 2013, via Planning Commission). Sustainable yield is lower than total recharge, so the true overstress is larger.
• The Barind Tract water table falls about 0.30 m per year on average, 1 to 2 m in Nachole and Tanore, with a cumulative drop near 15 m (BWDB/BMDA). Rainfall of 1,200 to 1,400 mm cannot recharge in-season pumping.
• Dhaka relies heavily on groundwater from the Dupi Tila aquifer (DWASA), a single-point-of-failure with no short-run substitute online.
• 10.6% of the population drinks household water above the 50 ppb arsenic standard (Bangladesh MICS 2019, DPHE/BBS/UNICEF), down from 12.5% of household source water in 2012-13; about 20 million are exposed above the standard (DPHE Situation Analysis 2009) and roughly 75 million at risk across 59 of 64 districts (CPD).
• The Bangladesh Water Act 2013 requires permits to withdraw water but sets no maximum withdrawal, so the legal cap on over-extraction does not yet bind (Planning Commission).

The extraction-recharge imbalance is structural, not cyclical

Bangladesh runs a food system on a multi-decade mining debt. Most irrigation comes from groundwater, drawn through millions of tubewells, the majority of them shallow (BADC/DPHE). Boro rice, irrigated December to May, takes the bulk of its water from aquifers (BRRI/BADC). That model multiplied rice output since the 1960s, but at about 30.2 BCM extracted against roughly 21.2 BCM recharged, the system loses near 9 BCM each year (MPO 1997; FAO 2013, via Planning Commission). Because sustainable yield sits below total recharge, the deficit understates the strain on the resource.

The deficit is not spread evenly. The Barind Tract and Dhaka face depletion; southern coastal aquifers face salinity intrusion; eastern floodplains face arsenic. A single national average masks three distinct failure modes, and a uniform subsidy regime treats them as one.

The price signal is absent. Subsidized electricity and diesel for irrigation pumps remove any incentive for efficient use, and the Water Act 2013 permit requirement carries no withdrawal ceiling (Planning Commission). Without extraction caps tied to aquifer levels, the overdraft widens every boro season.

The Barind Tract and Dhaka are the depletion frontiers

The Barind Tract in Rajshahi division is the most acute zone. Rainfall of 1,200 to 1,400 mm against a national mean near 2,300 mm, porous laterite soils, and intensive boro cultivation have pulled the water table down about 15 m cumulatively at roughly 0.30 m per year, with Nachole and Tanore losing 1 to 2 m per year (BWDB/BMDA). When shallow tubewells fail, boro production in the northwest is at risk, and the country's poorest districts face food insecurity and out-migration. BMDA's managed aquifer recharge, using check dams and recharge wells, shows localized gains but covers only a fraction of the stressed zone.

Dhaka is the second frontier. DWASA draws the majority of the capital's supply from the Dupi Tila aquifer (DWASA), with central water tables in long decline. The Padma surface-water plant is the planned alternative, but commissioning and distribution integration remain incomplete. A megacity on a compacting aquifer is a structural risk with no short-run exit, because aquifer compaction, once it occurs, is irreversible.

The arsenic crisis is receding in wells but not in exposure

10.6% of the population now drinks household water above the 50 ppb national standard, down from 12.5% of household source water in 2012-13 as deep tubewells displaced shallow ones (Bangladesh MICS 2019, DPHE/BBS/UNICEF). The improvement in tested wells does not clear the exposure: people in 59 of 64 districts still drink arsenic-affected water, with about 20 million exposed above the standard (DPHE Situation Analysis 2009) and roughly 75 million at risk (CPD). What Smith, Lingas and Rahman called the largest mass poisoning of a population in history (WHO Bulletin, 2000) carries a 10 to 20 year cancer latency, so the health burden of decades of contaminated water is still arriving.

The 50 ppb standard is itself a policy artifact: Bangladesh set it because retrofitting every well to the WHO 10 ppb guideline was fiscally costly. That pragmatic call has hardened into long acceptance of preventable harm, and the standard has not moved in two decades.

Recommendations

1. WARPO and BWDB should make the Water Act 2013 permit a binding cap, starting in the Barind Tract and Dhaka. The Act requires a permit but sets no maximum withdrawal (Planning Commission); the reform is to attach district extraction ceilings tied to real-time piezometric data and stand up telemetric water-level and quality sensors across the 64 districts within 36 months. Success signal: every Barind and Dhaka district has a published, enforced annual extraction ceiling and live piezometric telemetry by FY2029. Without an enforceable cap, recharge and efficiency investments are erased by unregulated pumping.
2. SREDA and IDCOL should meter every solar irrigation pump and price abstraction above an aquifer-safe block. The diesel-to-solar conversion cuts fuel cost and carbon but pushes the marginal cost of pumping toward zero; IWMI warns this can raise total abstraction without metering and limits. Fit flow meters at installation and charge for volume drawn above a sustainable threshold. Success signal: 100% of newly installed solar pumps are metered and on volumetric tariffs, with measured per-pump abstraction flat or falling year on year.
3. DPHE should commit to the WHO 10 ppb arsenic standard by 2030 with a phased pipeline. Priority high-exceedance districts need piped surface water for larger settlements and arsenic-monitored deep tubewells for smaller clusters. Testing should reach 100% of public and community tubewells, with results in a public national database. Success signal: the share of population drinking household water above 50 ppb falls below 5% (from 10.6% in MICS 2019) and the national tubewell test register is published and current.
4. BMDA should scale managed aquifer recharge across the full Barind stressed zone, funded from upper-block irrigation tariffs. Localized check dams and recharge wells already work; revenue from the priced extraction blocks above should fund expansion to the whole tract, plus alternate wetting and drying and crop diversification toward pulses and oilseeds that cut boro water demand. Success signal: managed recharge coverage and AWD adoption rise to a majority of Barind boro area, with the Barind decline rate slowing below 0.30 m per year.

What would change this view

The depletion conclusion rests on extraction exceeding recharge year after year. It would weaken if BWDB's next national assessment, built on the wider piezometric and GRACE record rather than the legacy MPO/FAO recharge estimate, showed recharge revised materially upward or rising toward extraction through wetter monsoons or large-scale managed recharge. The arsenic priority would narrow if a national re-survey on the MICS basis found household exceedance well below 10.6% and concentrated in a few districts. The capital's single-point-of-failure risk would ease if the Padma surface-water supply reached full Dhaka commissioning ahead of schedule. None of these is on the current trajectory, but each is the specific evidence that would revise the conclusion.