Production and Availability Forecasting

Production and Availability Forecasting

A borewell production forecast estimates the water that a defined operating arrangement can deliver during a stated period. It combines a permitted duty with a schedule, expected interruptions and a clear volume boundary. It must not turn an observed test rate into an unconditional monthly supply promise. The forecast should remain traceable to the operating decision and change when the underlying conditions change.

The most useful forecast separates what the well may produce, what the equipment can deliver, what the delivery system retains and what the user actually needs. These quantities can differ. A tank can conceal an hourly mismatch but cannot create water across a sustained supply deficit. An availability assumption can represent expected downtime, but it cannot establish that groundwater conditions support the remaining operating hours.

Define the forecast question

Start with the decision the forecast will support. A monthly procurement plan needs a volume estimate and uncertainty range. A daily production plan also needs the timing of delivery and usable storage. A contingency plan needs an explicit interruption scenario. Combining all three into a single “monthly capacity” number hides the period in which service may fail.

Record the start and end of the forecast, calendar convention, working days and operating shifts. Include planned shutdowns and changes in demand. In a multiwell system, identify whether simultaneous operation is permitted and supported by evidence. Summing standalone well rates can overstate deliverable supply if the combined operating arrangement has not been assessed.

Choose the volume boundary

Specify whether the forecast refers to discharged water at the wellhead, water entering a storage tank or usable water delivered to the customer process. A loss between two boundaries must be measured or explicitly assumed. Do not label a planning allowance as an observed loss. Where treatment rejects, backwashing or transfer losses matter, represent them with separate terms so their effect can be revised independently.

Use the same boundary when comparing forecast and actual. A wellhead forecast compared with a downstream process meter can show an apparent production shortfall even when the well performs as planned. The mismatch may be in transfer, storage or demand accounting.

Build the deterministic calculation first

For a constant approved rate Q, scheduled pumping hours H and an availability factor A defined against those scheduled hours, forecast discharge is V = Q × H × A. Use Q in m³/h and H in hours to obtain m³. A is dimensionless and must have a stated denominator. If planned outages have already been removed from H, do not deduct them again through A.

Where rate or schedule changes, calculate each interval separately: V = sum of Q_i × H_i × A_i. Keep the interval definitions in the worksheet. A single blended rate can be useful for presentation, but preserve the underlying operating states. This is particularly important when a restriction applies only during part of the month or when alternate wells operate on different schedules.

Separate availability from utilisation

Availability describes the share of a defined period during which the equipment or system is able to perform its required function under the chosen definition. Utilisation describes how much it is actually called upon to operate. A healthy well held off because its tank is full is not necessarily unavailable. A pump that cannot run because of a fault is a different state.

Define planned and unplanned interruptions consistently. A planned maintenance shutdown can be removed from the schedule or included in an availability model, but not both. Record whether grid interruptions, staffing constraints and downstream restrictions are included. Different definitions can produce different percentages from the same event log; the forecast must disclose its own definition rather than assuming the word “availability” is self-explanatory.

Worked monthly forecast

Consider a fictional approved planning duty of 10 m³/h for eight scheduled hours on 25 days. Scheduled hours are 200. At an illustrative availability assumption of 0.95, forecast wellhead discharge is 10 × 200 × 0.95 = 1,900 m³. This result is an arithmetic scenario. It is not a measured monthly result or evidence that the duty is sustainable.

Suppose four of the 200 scheduled hours are already removed for planned maintenance. The revised schedule is 196 hours. If the 0.95 factor represents only additional unplanned downtime, the forecast becomes 1,862 m³. If that factor already included the same four maintenance hours, the calculation would double-count the planned interruption. Retain a note identifying exactly which events each term represents.

Model demand timing and storage

A monthly volume balance is necessary but insufficient for a reliable supply plan. Compare delivery and consumption over the intervals that matter to the operation. Track usable storage at the beginning and end of each interval. Storage change equals inflow minus outflow within the stated boundary, adjusted for measured or assumed losses. Keep minimum usable storage distinct from the tank’s geometric capacity.

For an illustrative tank balance, an opening usable volume of 30 m³, incoming volume of 50 m³ and demand of 65 m³ produces a closing usable volume of 15 m³ before other terms. If the operating reserve is 20 m³, the plan breaches its reserve even though the tank does not become empty. The reserve itself must come from the service plan; the example does not prescribe a universal reserve.

Use scenarios that expose the real uncertainty

A base case should describe the current operating basis. A constrained case should change identifiable conditions such as fewer permitted hours, lower verified rate or longer interruption. An improved case should be tied to an achievable change, not a desired target. Label each assumption and show which evidence could confirm or reject it.

Avoid multiplying several optimistic assumptions together while reporting only the final volume. A higher rate, longer schedule and better availability may each need separate approval or investment. If one is unresolved, show the effect explicitly. The forecast should help the owner choose an action, such as additional storage or an alternative supply arrangement, rather than merely displaying a larger theoretical number.

Treat seasonality as evidence rather than a percentage habit

A seasonal adjustment should arise from relevant observations or a clearly stated planning assumption. Do not apply a universal dry-season factor to every well. Identify whether the record spans the limiting period, whether the operating schedule changed and whether neighbouring conditions are comparable. A short record may justify a broader uncertainty statement rather than a precise adjustment.

Keep the date of the operating approval visible in the forecast. If a seasonal observation invalidates the approved duty, revise that duty through its control process before presenting a new forecast as authorised supply. The forecasting worksheet should consume the operating decision; it should not silently create a new operating permission by changing a cell.

Check uncertainty without inventing confidence

Separate measurement uncertainty from scenario uncertainty. Flow measurement error affects the estimate of delivered volume. Unknown future outages affect the operating schedule. Uncertain groundwater response affects whether the proposed duty remains appropriate. These are not interchangeable, and combining them into one unexplained percentage can obscure the action needed to reduce uncertainty.

Use ranges only when their basis is stated. A low and high planning scenario is not automatically a statistical confidence interval. If a probability claim is made, the model, data and assumptions supporting it must be documented. For many operating decisions, a transparent set of bounded scenarios is more useful than a precise-looking probability unsupported by the available record.

Turn the forecast into an accountable operating plan

Assign an owner to each material assumption. Operations can confirm the schedule, maintenance can supply planned interruptions, the responsible technical reviewer can confirm the permitted duty and the demand owner can validate consumption. Record the date of each input. This makes it possible to distinguish a calculation error from a forecast that became outdated because the business changed.

Freeze an issued forecast before the period starts. Preserve later revisions with reasons and effective dates. At period end, compare actual results against the version that was used to make the decision, not a revised forecast adjusted to resemble the outcome.

Construct an auditable forecasting worksheet

Arrange the worksheet so that inputs, calculations and outputs are distinguishable. Inputs should include the approved duty reference, schedule, interruption assumptions, demand and storage basis. Calculations should show the interval arithmetic. Outputs should identify forecast discharge, forecast delivery and any service shortfall. Avoid hiding important assumptions in cell comments that disappear during export.

Give each input a status such as measured, approved, assumed or proposed, using definitions in the worksheet notes. Record its source and date. This allows the reviewer to identify which uncertainty can be resolved through measurement and which depends on a future decision. A proposed higher rate must not be mixed with approved hours and presented as an approved base case.

Protect the issued forecast from silent alteration while keeping a working copy available for revision. Record the issue date and forecast period. At the end of the period, retain the inputs that were actually used. This is necessary to learn whether the forecast was wrong because of its model, its assumptions or a later change in circumstances.

Model partial-month restrictions explicitly

If the duty changes during a month, divide the forecast into the applicable periods. Do not average the old and new rates without weighting their scheduled hours. Record the effective date of the restriction and whether already completed production is included as actual or remains part of a forecast view.

For an illustrative month, ten days at 12 m³/h for six hours give 720 m³. Fifteen days at 9 m³/h for six hours give 810 m³. Before availability adjustments, the combined forecast is 1,530 m³. A simple average rate of 10.5 m³/h applied to all 150 hours would give 1,575 m³ and would not represent the stated schedule.

If the restriction’s end date is uncertain, show scenarios instead of inventing one definitive return date. A management decision can then consider the supply consequence of a longer restriction. The forecast should expose dependence on restoration approval rather than conceal that dependence inside a blended monthly number.

Reconcile storage without double counting production

Water entering storage is not new production when it later leaves the tank. Keep the wellhead, storage and consumption balances distinct. Record opening and closing usable storage for the reporting period, as well as transfers to or from other supplies. Otherwise, a period that starts with a full tank can appear to deliver more water than the well produced without an explanation.

For a fictional daily balance, the well supplies 80 m³, an alternative source supplies 10 m³ and usable storage falls from 25 to 15 m³. The available volume for consumption and other outflows is 100 m³ before separately accounted losses. The extra 10 m³ from storage is an inventory reduction, not an improvement in well yield.

A forecast should identify how long such inventory support can continue. Repeatedly drawing down storage cannot sustain a permanent deficit. Show the lowest expected usable storage and the trigger for the alternative supply plan. The storage calculation should use the actual usable range, not a nominal tank volume that includes inaccessible or reserved capacity.

Avoid unsupported probability language

A planning scenario labelled conservative is not automatically a 95% confidence forecast. Probability language requires a defined statistical or probabilistic basis. If the inputs are expert assumptions with limited historical data, describe them as assumptions and scenarios. This remains useful for decision-making without suggesting a level of statistical support that does not exist.

Where historical downtime is used, examine whether the future operating arrangement is comparable. New equipment, changed maintenance, different power conditions or a revised duty can alter the relevance of the record. Do not carry forward a favourable average simply because it improves the forecast. Preserve the evidence for choosing the planning input.

Use sensitivity to identify the dominant uncertainty. Show how the result changes when one material input changes while others remain fixed. This is a diagnostic tool, not a prediction that only one input will change in reality. The most useful outcome may be a decision to improve a measurement or secure a contingency rather than refine a decimal place.

Forecast a multiwell arrangement carefully

List the duty and schedule of each well separately, then identify the assumptions about simultaneous operation and shared infrastructure. A common electrical supply, discharge main or storage tank can create a shared constraint. The combined forecast should not exceed a downstream limit that the individual well worksheets do not represent.

Where interaction between wells is relevant, use reviewed evidence for the proposed combined duty. Standalone observations do not automatically establish the same performance when several wells operate together. Record which combinations were observed and which remain proposed. The forecasting model should consume that technical conclusion, not create it by summing rates.

Track allocation as well as total output. If one well is restricted, the plan may shift hours to another. That shift requires checking the receiving well’s approved conditions and the shared system. A simple total-volume target must not cause an unreviewed redistribution of duty across assets.

Use the forecast to choose a practical contingency

Define the consequence of the constrained scenario: delayed production, reduced service, additional storage drawdown or alternative supply. Compare responses using their actual availability, timing and operating implications. Do not assume that a backup source can deliver instantly or indefinitely. Its own boundary and evidence should be documented.

The contingency trigger should occur early enough for the selected response to be useful. A forecast can identify when storage or delivery is likely to become inadequate, but the trigger must be tied to observable information and responsible action. Avoid a plan that waits until the tank is empty to discover whether an alternative supply exists.

Record who owns the contingency decision and how it affects the normal forecast. Once an alternative supply is used, keep its volume separate from borewell production. This preserves the ability to evaluate the well and the service plan independently. Reliable customer delivery can coexist with reduced well production when the alternative contribution is transparently accounted for.

Forecast review questions before issue

Check whether every rate is linked to a current decision, every hour belongs to a defined schedule and every reduction factor has a clear denominator. Confirm that planned interruptions are not counted twice and that a proposed restoration is not presented as already approved. Review the volume boundary against the meter that will later supply actual results.

Then examine the consequence of being wrong. If the service impact is material, confirm that the contingency and monitoring arrangements are adequate for the uncertainty. Forecast review should end with a usable decision: issue the base case, retain a stated condition, obtain specific evidence or revise the plan. It should not end with an unexplained number that has no accountable owner.

Questions about borewell forecasts

Can a test rate be multiplied by every hour in a month

Only as a clearly labelled arithmetic maximum under that assumption, not as a justified operating forecast. The approved schedule, observed response, equipment limits and interruption assumptions must support the duty used for planning.

Does a 95% availability factor guarantee delivery

No. It is a model input whose meaning depends on its denominator and included events. It does not guarantee future performance or remove groundwater and demand uncertainty.

Can storage correct a permanent volume deficit

Storage can shift delivery in time. It cannot indefinitely compensate when total available inflow remains below total demand. Evaluate both the interval balance and the period balance.

What should accompany a forecast request

Supply the current operating instruction, rate and runtime records, totaliser information, planned shutdowns, demand profile, usable storage and required service continuity. Identify which values are measured, approved or assumed so the resulting forecast remains interpretable.

Source notes

The benchmark attachment is a directory, not a recovered article body or measurement dataset.

Content-structure method attribution: Eugen Ullrich, eullrich.com, GEO Content Engineering, CC BY 4.0. No ranking or AI citation guarantee is implied.