1.5 Raw Water Storage, Release Scheduling & Demand Forecasting
Key Takeaways
- Maximum day demand typically runs 1.5 to 3.0 times average day demand, and peak hour demand typically runs 3 to 6 times average day, so plants and storage must be sized and operated against peaks rather than averages.
- Total storage serves three separable functions — equalizing (diurnal) storage, fire flow storage, and emergency reserve — and an operator must know which portion of the tank is actually available before drawing it down.
- Safe yield is the dependable withdrawal a reservoir can sustain through the design drought; withdrawing above safe yield during a drawdown period borrows from future supply rather than creating new water.
- Release scheduling must satisfy downstream obligations such as minimum instream flow, compensation releases, and permit conditions, which are legal constraints that exist regardless of the utility's own demand.
- Excess finished water storage raises water age, which drives chlorine residual loss, nitrification in chloraminated systems, and disinfection byproduct formation, so turnover is an operating target and not merely a design concern.
Why Demand Forecasting Is an Operator Task
A treatment plant does not get to choose when customers use water. Demand arrives on a daily cycle, a weekly cycle, and a seasonal cycle, and the operator's job is to have treated water in storage before the peak, using a production rate the plant can actually sustain at compliant quality. Producing reactively — chasing a falling tank level by ramping the plant hard — is how operators blow through filter loading limits, shorten contact time, and turn a demand event into a turbidity or CT violation.
The ABC Class II criteria state this directly as determine and adjust plant flows to meet system demands and manage stored water release based on forecasted demand.
Demand Patterns and Peaking Factors
Base Terminology
- Average Day Demand (ADD): Total annual production divided by 365.
- Maximum Day Demand (MDD): The highest single-day production in the year. This is normally the design basis for treatment plant capacity.
- Peak Hour Demand (PHD): The highest one-hour rate in the year. This is normally the design basis for distribution piping and high-service pumping, absorbed largely by storage.
- Per Capita Demand: Commonly expressed in gallons per capita per day (gpcd). Residential use often falls near 80 to 100 gpcd, with total system per-capita figures running higher where commercial or industrial load is significant.
Typical Peaking Factors
| Ratio | Typical Range | What It Drives |
|---|---|---|
| MDD / ADD | 1.5 – 3.0 | Treatment plant rated capacity, source capacity |
| PHD / ADD | 3.0 – 6.0 | Transmission mains, high-service pumps, equalizing storage |
| PHD / MDD | 1.5 – 2.0 | Short-term storage drawdown rate |
Small systems and heavily residential systems sit at the high end of these ranges because their demand is less diversified: everyone showers, irrigates, and cooks at roughly the same time. Large systems with substantial industrial baseload sit at the low end.
The Diurnal Curve
A typical residential system shows two daily peaks — a sharp morning peak roughly between 6 and 9 a.m. and a broader evening peak roughly between 5 and 9 p.m. — separated by a midday plateau and a deep overnight minimum between about midnight and 5 a.m.
Operators exploit the overnight minimum deliberately. Running the plant and high-service pumps hard overnight refills storage, buys off-peak electric rates, and leaves the plant with headroom during the day. This is the practical link between demand forecasting and the energy optimization task.
Seasonal and Weather Drivers
- Irrigation is usually the largest seasonal swing. A hot, dry July can double summer demand relative to a wet spring.
- Temperature and rainfall forecasts are leading indicators; a forecast of several consecutive days above 90°F with no rain warrants pre-filling storage.
- Freeze events raise demand through customer drip lines and main breaks, and simultaneously reduce plant capacity through cold-water coagulation and longer required CT.
- Local events — a holiday weekend, a fair, a school calendar, a large industrial customer's production schedule — produce predictable single-day deviations.
The Three Functional Components of Storage
Total tank volume is not all available for routine operation. Storage is conventionally divided into three stacked components:
| Component | Purpose | Operating Status |
|---|---|---|
| Equalizing (operating) storage | Absorbs the difference between steady plant production and the varying diurnal demand | Actively cycled every day |
| Fire flow storage | Supplies the required fire flow rate for the required duration as set by the fire authority or insurance rating | Reserved; not drawn for routine demand |
| Emergency reserve | Covers source loss, power failure, main break, or plant outage | Reserved; drawn only under contingency |
An operator who draws a tank down into the fire reserve to ride out an ordinary summer afternoon has removed the system's fire protection without telling anyone. Knowing the elevation that separates equalizing storage from fire reserve is a basic operating fact, and it belongs on the SCADA display as a defined setpoint.
Storage Level Setpoints
Practical plant control ties production to tank level through documented setpoints: a refill-start level that brings additional filters, wells, or high-service pumps online, a refill-stop level near overflow that prevents spilling, and a low-low alarm that signals the equalizing volume is exhausted. Automating these in SCADA is common; knowing what they mean, and what to do when they are reached during a communications failure, is the operator's responsibility.
Raw Water Reservoir Management
Yield Concepts
- Safe yield (also called dependable or firm yield): the withdrawal rate the reservoir can sustain through the worst drought of record used for design. This is the honest long-run capacity number.
- Drawdown: the decline in reservoir surface elevation as withdrawal exceeds inflow. A reservoir operated below safe yield during a dry period is consuming carryover storage.
- Carryover storage: volume intentionally held over from wet years to cover multi-year droughts.
Withdrawing above safe yield does not create water. It moves supply forward in time, and it must be paired with an explicit plan for how the deficit will be recovered or how demand will be curtailed.
Storage Release Scheduling
Reservoir releases are rarely discretionary. Typical obligations include:
- Minimum instream flow required by permit to protect downstream aquatic habitat.
- Compensation releases owed to downstream riparian or senior water rights holders.
- Flood control drawdown required seasonally by the controlling authority so the reservoir has room to absorb a design storm.
- Reservoir-to-reservoir transfers where a utility operates a storage impoundment upstream of the treatment intake and must release water in advance of the demand it will serve, accounting for channel travel time.
Travel time is the operational catch. If the release point is a day's travel upstream of the intake, the release decision must be made a day ahead of the demand it serves, which is precisely why the forecast matters.
Selective Withdrawal Coordination
Release and withdrawal decisions interact with the limnology covered earlier. Drawing down a stratified reservoir late in the summer can lower the intake's effective position relative to the thermocline, pulling anoxic hypolimnetic water — high in soluble iron, manganese, ammonia, and hydrogen sulfide — into the plant. A drawdown schedule should be checked against the multi-level intake's available withdrawal ports.
Drought Contingency Staging
Most utilities operate a staged drought plan keyed to reservoir elevation, streamflow, or groundwater level. A representative structure:
| Stage | Typical Trigger | Typical Action |
|---|---|---|
| Stage 1 — Advisory | Storage below a defined seasonal target | Voluntary conservation messaging; increase monitoring frequency |
| Stage 2 — Warning | Continued decline; supply projected short within months | Mandatory odd/even or day-of-week irrigation restrictions |
| Stage 3 — Emergency | Supply projected short within weeks | Ban on outdoor irrigation, vehicle washing, and filling pools |
| Stage 4 — Critical | Imminent inability to meet essential demand | Essential-use only; activate interconnections and hauling |
Restrictions work primarily by removing discretionary outdoor use, which is why they are aimed at irrigation first. That is also where the demand elasticity is.
Water Age: The Cost of Carrying Too Much Storage
More finished storage is not automatically better. Volume held without turnover increases water age, and water age degrades quality in several linked ways:
- Disinfectant residual decay. Free chlorine and monochloramine both decay with time; long detention in a tank can drive the residual below the minimum required in the distribution system.
- Nitrification in chloraminated systems. As monochloramine decays it releases free ammonia, which nitrifying bacteria oxidize to nitrite and nitrate. Nitrification consumes residual further, depresses pH, and elevates heterotrophic plate counts. Warm, stagnant tank water is the classic incubator.
- Disinfection byproduct formation. TTHM formation continues as long as free chlorine and precursor organics are in contact, so the oldest water in the system frequently produces the highest locational running annual average.
- Thermal stratification inside the tank, which effectively shrinks the mixed volume and makes the aged fraction worse than the average detention time suggests.
The operating countermeasure is deliberate turnover: cycling tanks through a meaningful daily level change rather than floating them near full, and in some cases installing active mixing systems. A common operating target is to turn over a substantial fraction of the tank volume each day, with the specific target set by the system's residual and byproduct performance.
Putting It Together: A Production Decision
Given a forecast of 95°F with no rain, a current usable equalizing volume of 0.6 MG, an average day demand of 2.0 MGD, an expected peaking factor near the high end at 2.5, and a plant rated at 4.0 MGD, the operator should recognize that expected demand near 5.0 MGD exceeds plant capacity. The correct response is to fill storage overnight at the plant's sustainable rate, plan to cover the daytime deficit from equalizing storage, verify that the drawdown will not enter fire reserve, and, if it will, initiate the conservation stage rather than overdriving the filters.
A residential water system has an average day demand of 1.8 MGD. During the summer peak, which single-day production figure is most consistent with typical maximum day peaking factors for a small, heavily residential system?
During an afternoon demand peak, a finished water tank level falls below the elevation that the utility has designated as the top of its fire flow reserve. What is the significance of that elevation?
A chloraminated system has recently added finished water storage and now floats its tanks near full, achieving very little daily level change. Over the following weeks operators observe declining monochloramine residual, rising nitrite, and elevated heterotrophic plate counts in the affected pressure zone. What is the most likely cause?