ESRS E3 Water: Withdrawal, Consumption, Discharge and Water-Stress Reporting

Educational material. It does not replace the applicable delegated act, national law, legal advice or an assurance conclusion.

A cubic metre has different sustainability implications depending on where and when it is withdrawn or consumed. One site may operate in a water-abundant basin with strong infrastructure, while another draws from a stressed catchment shared with communities and ecosystems. A group total is therefore necessary but not sufficient for understanding impacts, dependencies and risks.

Revised E3 links own-operation metrics to the broader ESRS architecture. The undertaking identifies material water and marine-resource IROs, describes policies, actions and targets, explains methods and reports financial effects through ESRS 2 SBM-3. The metric design should be driven by those material IROs and should preserve enough site or basin granularity to explain them.

Source anchor: revised ESRS E3, paragraphs 1-6 and 16; AR 1-5; ESRS 2 GDR-P/A/M/T and SBM-3.

The site master is the control population for E3. It should include legal entity, site name, coordinates, ownership or operational-control status, activities, reporting period, water sources, discharge destinations, storage, meter population, permits and basin identifiers. A group should not rely only on a utility-account list because privately abstracted water, shared facilities, temporary sites and acquired operations may be missing.

Data should normally be collected at the level at which the physical flow can be measured and the basin can be identified. Several meters may feed one site balance, and several sites may lie in one basin. Central consolidation should preserve the site and basin keys even if the public disclosure presents an aggregated group metric. This makes it possible to disclose or explain material concentrations without rebuilding the dataset.

Water withdrawal records the volume entering the operational boundary from external sources. The source classification should be consistent across sites and years. Typical categories include surface water, groundwater, seawater, produced water and third-party water such as municipal supply. Where rainwater is collected, the methodology should state whether and how it is included.

The preferred source is a calibrated meter covering the reporting period. Utility invoices can be strong evidence when the billing period and site match, but invoices may include estimates or span year-end. Private abstraction often requires meter readings and permit returns. When a site has no complete meter, a controlled calculation may use equipment flow, operating hours, production ratios or a proxy site, with an uncertainty and improvement plan.

Reconcile opening and closing meter readings to the period and investigate negative or implausible movements.

Separate shared meters using a documented allocation driver and review it annually.

Identify purchased water contained in raw materials or products only where it meets the metric definition and is material.

Do not add internal recycled or reused water to withdrawal; it is not a new external inflow.

Flag acquisitions, disposals, shutdowns and unusual production so variance analysis explains real operational change.

Water discharge is the volume leaving the undertaking's own operations. The destination matters because a direct release to a river, a transfer to sewer, a discharge to a third-party treatment plant and water supplied to another user have different environmental and risk implications. The reporting method should identify whether the volume is measured at the outfall, estimated from inflow, calculated from production or derived from a permit return.

Discharge quantity is not a water-quality metric. Where water-pollution impacts are material, pollutant loads are reported under E2 and connected to the E3 flow and receiving context. A large clean cooling-water discharge and a smaller contaminated process-water discharge should not be interpreted solely by volume.

Revised E3 AR 4 expresses consumption as C = W - D. This simple equation is the starting point, but the undertaking should understand the physical reasons for the difference. Water may evaporate, be incorporated into product, be consumed by people or animals, leave in waste or sludge, or remain in storage. A negative consumption result normally indicates a boundary, cut-off, meter or storage problem that needs investigation.

If a site has material reservoirs, tanks, tailings facilities, ponds or process inventory, opening and closing storage can affect the relationship between annual withdrawal and discharge. The organisation should define whether E3's stored-water metric is a closing balance, average or other prescribed basis and keep it separate from annual flows. Where a storage adjustment is needed to explain consumption, the calculation and rationale should be documented.

Recycled or reused water can reduce the need for external withdrawal, but it is not additional withdrawal. A cubic metre recirculated five times is still one external cubic metre entering the system. The reporting model should therefore record internal recycle loops separately. The flow can support efficiency analysis and targets without inflating the physical water balance.

Figure 1. ESRS E3 water balance with a water-stress overlay. Original LRA practitioner visual.

Water stress is generally a relationship between water demand and available renewable supply. It is not identical to drought, flood, water quality, affordability or access. A site may face low basin stress but high flood risk, or high stress while having secure contracted supply. The E3 assessment should therefore use stress as one part of the location context and consider other material impacts and dependencies separately.

The undertaking should document the basin dataset, version, geographic resolution, stress threshold, reporting year and treatment of sites near basin boundaries. It should apply the method consistently and review changes when a dataset is updated. If two reputable tools produce different classifications, the organisation should investigate rather than choosing the more favourable result.

Global datasets are screening tools. Local evidence may include abstraction restrictions, drought orders, community access issues, ecological status, seasonal scarcity, groundwater depletion, permit constraints and infrastructure reliability. A site classified as low stress should not be assumed immaterial where local evidence indicates a serious impact or dependency.

The public disclosure can present group totals, but material site or basin concentrations should be explained where necessary to understand the IRO. A percentage of total consumption in high-stress areas is useful, but it should not hide that one site accounts for most of the exposure or that seasonal consumption is concentrated in the dry period.

ESRS does not require every pipe to be metered before reporting can begin. It does require the method and significant assumptions to be understandable. The estimate hierarchy should prioritise metered data, utility records, permitted abstraction/discharge returns, equipment flow and operating hours, production-based models and proxy-site estimates. The reporting team should quantify the share of the group metric that is estimated where this is material.

A useful water target is linked to the material IRO and its location. An absolute group reduction may be less decision-useful than a site-level consumption target in a stressed basin, a leakage target, a reuse target or an ecological flow commitment. The target should specify the boundary, baseline, target year, metric, stress-area relevance, assumptions and whether increased production changes the interpretation.

Actions can include process redesign, closed-loop cooling, leakage reduction, alternative sourcing, treatment and reuse, nature-based catchment measures, supplier engagement and community water-access projects. The evidence should distinguish an action completed from an environmental outcome achieved. Installing a recycling system is not the same as demonstrating reduced external withdrawal or improved basin condition.

Current and anticipated financial effects may include water purchase and treatment cost, capex, production interruption, permit constraints, asset impairment, insurance, supply-chain disruption and revenue from water-efficient products. Finance should map each material E3 IRO to recognised amounts and anticipated effects without presenting qualitative risk as an already recognised accounting provision.

Figure 2. ESRS E3 evidence chain from site to sustainability statement. Original LRA practitioner visual.

Cedar creates a geocoded site master and maps each plant to a basin using a controlled dataset. Local review confirms seasonal restrictions at one plant that the global dataset classifies as medium rather than high stress. The materiality assessment treats the plant as a priority because community water access and production continuity are affected.

The group records 1,000,000 m3 of withdrawal and 760,000 m3 of discharge, producing 240,000 m3 of consumption. It records 400,000 m3 of internal reuse as a separate memo flow. The reservoir increase is analysed so the period balance is not misinterpreted. The shared-meter plant is allocated using machine operating hours, supported by a three-month sub-meter study; 8% of the group withdrawal remains estimated and is included in the improvement plan.

Of total consumption, 90,000 m3 occurs in high-stress areas. Cedar sets a site-specific target to reduce consumption at those plants by 20% by 2030, while a group-wide reuse target supports the action plan. Finance links the material risk to expected capex, possible production interruption and increased water-treatment cost.

Why it is stronger: the wording states the balance, separates reuse, identifies high-stress subsets, names the method and local validation, quantifies estimated data and describes the improvement action. A complete disclosure would also provide required policies, actions, targets and financial-effects information linked to material IROs.