GPopulation Forecasting Methods for Water Demand
The design population sets the size of the whole scheme. Here is how the standard forecasting methods differ and how to pick the one that fits your settlement.
Why the forecast decides everything
A water supply scheme is sized for the population at the end of its design period, not for today. Every downstream figure — source yield, pump capacity, main diameter and storage volume — scales directly with that forecast population, so an error here is multiplied through the entire design.
The aim is a realistic, defensible projection: optimistic enough that the scheme is not inadequate on the day it is commissioned, but not so inflated that money is sunk into capacity that is never used.
Arithmetic increase method
The arithmetic method assumes the population grows by a constant number of persons per decade — the average of past increases. You add that fixed increment for each future decade to reach the design population.
It suits large, old and essentially saturated towns whose growth has flattened. Applied to a fast-growing settlement it under-estimates badly, because real growth there compounds rather than adds.
Geometric increase method
The geometric method assumes a constant percentage growth rate, so the population compounds like interest. You apply the average past growth rate forward over the design period.
It fits young, rapidly expanding settlements, but over a long horizon it can over-estimate dramatically, because no real population sustains a high percentage rate indefinitely. It is best used over shorter periods or with a tapering rate.
Incremental increase method
The incremental increase method is a middle path: it takes the arithmetic increment and adjusts it by the average change in that increment from decade to decade. In effect it allows the rate of growth itself to rise or fall.
Because it captures acceleration or deceleration in growth, it often gives the most balanced forecast for ordinary towns and large villages, which is why many departmental norms lean towards it.
From population to design demand
Once the design population is fixed, multiply it by the prescribed per-capita demand for the service level to get the average daily demand, then apply the peak factors for maximum day and peak hour to size the source, mains and storage respectively.
Always record which method you used and why, with the census figures behind it. A forecast whose basis is written down is one that survives scrutiny; an unexplained number invites a query at every later stage.
Choosing a forecasting method that fits the habitation
The arithmetic, geometric and incremental-increase methods each embed a different assumption about growth. Arithmetic increase suits a mature, slow-growing village; geometric increase fits a young, fast-expanding one but over-predicts if applied blindly over a long horizon; the incremental-increase method tempers both and is often the safest default where several censuses are available. The honest approach is to compute all three and defend the one you adopt against the habitation's actual census trend.
Water demand then follows from the forecast population and the per-capita supply figure appropriate to the service level — public standposts, private connections with or without sewerage — plus allowances for institutions, livestock, floating population and unavoidable losses. A forecast population attached to the wrong per-capita figure produces a confident but wrong design flow.
Design to a stated horizon (commonly 30 years) and record the method, the base census and the per-capita basis in the design report so the assumptions can be revisited rather than reverse-engineered later.
Which population forecasting method should I use?
Compute arithmetic, geometric and incremental-increase projections, compare them against the habitation's census history, and adopt the one that best matches the observed trend — commonly the incremental-increase method.
What design period is used for rural water supply?
A design horizon of about 30 years is commonly adopted, so the population is forecast to that year before demand is worked out.
How is water demand derived from population?
Multiply the forecast population by the per-capita supply for the service level, then add institutional, livestock, floating-population and loss allowances.
Why does the per-capita figure matter so much?
The same population at a standpost service level and at a private-connection-with-sewerage level produces very different demands, so using the wrong per-capita basis skews the whole design.