Storage Reservoir Operations, Mass Curves and Sequent Peak Algorithm

Unit 03 — Reservoir Systems

Unit overview. Reservoirs redistribute water through time; they do not create water. Storage evolves from inflow, controlled releases, withdrawals, precipitation on the surface, evaporation, seepage and spills. A storage–elevation–area relation is needed for actual design, because surface area and spillway discharge are not constant. Operators face competing targets: retain water before droughts, reserve empty flood-control space before storms, satisfy downstream minimum flows and preserve dam safety. Sediment changes the geometry, sometimes reduces live storage and can interfere with intakes. Planning requires sequences of wet and dry periods rather than one convenient design month.

Field problem. A synthetic 30-day inflow sequence contains six high-flow days followed by a long dry spell. Operate the same reservoir under a high conservation target and a more aggressive pre-release rule. Compare the peak spill, unmet downstream demand, days below reserve storage and total evaporation. The second rule may reduce flood risk but worsen dry-season supply. Note how the conclusion depends on the quality and lead time of weather forecasts.

Four topics. Storage, Operation, Flood Control, Sedimentation.

Before starting: Sketch a control volume and record which terms describe depths, rates, stored volumes, concentrations, entitlements or monetary values. Identify the boundary, the timescale, the user groups affected and at least one observation whose absence would make a numerical conclusion unreliable. Explain why seasonal variability matters more than a single annual figure for some decisions.

9. Storage

Core concept. Active storage buffers supply across periods, dead storage sits below normal outlets and flood-control storage provides space for extreme inflows. These regions can change as sediments accumulate. Continuity in volume is the core test of a reservoir routing model.

Analysis workflow. For a synthetic 12-month inflow sequence, compute closing storage with explicit monthly evaporation and deliveries. Record any forced spill or unserved demand; neither can be hidden as negative storage.

Engineering interpretation. An apparent annual surplus can coexist with empty storage at the end of a critical month.

Evidence and measurement. Work with an illustrative upland watershed where the available record includes rainfall and gauge coverage. Begin by naming the controlled variable, the response variable and the geographic location of each observation. Check the date, measurement frequency, applicable units and any uncertainty statement. Where a record is missing or derived from a different location, describe the interpolation explicitly rather than silently assuming that it is representative. Choose a realistic time step. Storm peaks may need minutes while aquifer or water-allocation trends may be represented over months; coarse averaging can erase the limiting event.

From conservation to prediction. A transparent first calculation should reconcile a mass, energy or financial balance as appropriate to the physical mechanism. Convert input quantities into coherent dimensions and display intermediate steps, particularly when mixing depth, area, volume, discharge and time. If a relationship uses a calibrated coefficient, record the period and conditions that support it. Separate a model output from an observation. A discrepancy may reveal a missing process, incorrectly measured inputs, model-structure error or a unit conversion mistake.

Scenario comparison. Compare a baseline in which current practice continues against an alternative with a different capacity, demand, governance rule or environmental reserve. Where possible, change only one assumption at a time, then carry out a second experiment in which multiple uncertain inputs vary together. Explain what each comparison isolates and what interactions remain hidden. Specify an observable metric and the unit in which it would be reported. A useful metric permits comparisons across dates without changing the spatial boundary or its denominator.

Scope and limitations. Do not interpret a screening model as an approved infrastructure design. Confirm whether physical boundary conditions, environmental constraints, public-health requirements and downstream effects appear in the calculation. A conceptual model that omits return flows, aquifer exchange, water quality or operational failures may overestimate usable supply. Before using a numerical method, distinguish a conservation equation from an empirical relation and a legally mandated constraint. Only conservation identities hold without calibration.

Practice questions with answer guidance.

1. Explain: How does the concept of storage connect to both the quantity and timing of water availability in reservoir systems? Answer guidance: Identify the physically relevant boundary, state the controlling process described above and name one observation that could falsify the assumed behavior.
2. Quantify: Choose two contrasting data periods for the upland watershed, state all units and compute one meaningful normalized comparison; distinguish an absolute difference from a percentage change. Answer guidance: Report the numerator, denominator, observation period, units and rounding precision; do not manufacture a statistical confidence interval without an error model.
3. Diagnose: Suppose a simple model fits the first half of the observation period but fails after a large storm or a prolonged dry season. Which neglected process would you investigate first? Answer guidance: Look for a process whose magnitude or regime changes with the event, then specify a measurement that could discriminate it from an input error.
4. Decide: What evidence would justify changing a withdrawal rule, reservoir policy or monitoring plan? Answer guidance: Compare safety and minimum-service constraints first; then evaluate operating performance, maintenance cost, affected stakeholders and the reversibility of the decision.

Mini field task. Write a one-page technical memo on storage for the upland watershed. Include a schematic boundary, a data-quality table, one auditable calculation, a counter-scenario and a clear statement of what the evidence does not establish. Do not present synthetic numbers as observed field data.

10. Operation

Core concept. A rule curve is an operating target, not a guarantee. Pre-release can reduce potential spill but sacrifices stored water if the storm misses the watershed. Operating rules depend on forecast skill, downstream channel capacity and safety margins.

Analysis workflow. Run the same inflow sequence with conservation-focused and flood-space-focused release rules. Plot release, storage and unmet demand; quantify which benefit changes and by how much.

Engineering interpretation. Risk preference must be explained because a rule optimized for average delivered volume may be poor under rare hazardous floods.

Evidence and measurement. Work with an illustrative seasonal monsoon basin where the available record includes measured flow and stage. Begin by naming the controlled variable, the response variable and the geographic location of each observation. Check the date, measurement frequency, applicable units and any uncertainty statement. Where a record is missing or derived from a different location, describe the interpolation explicitly rather than silently assuming that it is representative. Keep the original measurements, assumptions, units and script version in a reproducible notebook. A repeatable result is necessary but insufficient evidence of a valid model.

From conservation to prediction. A transparent first calculation should reconcile a mass, energy or financial balance as appropriate to the physical mechanism. Convert input quantities into coherent dimensions and display intermediate steps, particularly when mixing depth, area, volume, discharge and time. If a relationship uses a calibrated coefficient, record the period and conditions that support it. Sketch a conceptual system with each inflow, outflow, storage and feedback represented. A number without a transparent system boundary is not an auditable engineering result.

Scenario comparison. Compare a baseline in which current practice continues against an alternative with a different capacity, demand, governance rule or environmental reserve. Where possible, change only one assumption at a time, then carry out a second experiment in which multiple uncertain inputs vary together. Explain what each comparison isolates and what interactions remain hidden. Choose a realistic time step. Storm peaks may need minutes while aquifer or water-allocation trends may be represented over months; coarse averaging can erase the limiting event.

Scope and limitations. Do not interpret a screening model as an approved infrastructure design. Confirm whether physical boundary conditions, environmental constraints, public-health requirements and downstream effects appear in the calculation. A conceptual model that omits return flows, aquifer exchange, water quality or operational failures may overestimate usable supply. Separate a model output from an observation. A discrepancy may reveal a missing process, incorrectly measured inputs, model-structure error or a unit conversion mistake.

Practice questions with answer guidance.

1. Explain: How does the concept of operation connect to both the quantity and timing of water availability in reservoir systems? Answer guidance: Identify the physically relevant boundary, state the controlling process described above and name one observation that could falsify the assumed behavior.
2. Quantify: Choose two contrasting data periods for the seasonal monsoon basin, state all units and compute one meaningful normalized comparison; distinguish an absolute difference from a percentage change. Answer guidance: Report the numerator, denominator, observation period, units and rounding precision; do not manufacture a statistical confidence interval without an error model.
3. Diagnose: Suppose a simple model fits the first half of the observation period but fails after a large storm or a prolonged dry season. Which neglected process would you investigate first? Answer guidance: Look for a process whose magnitude or regime changes with the event, then specify a measurement that could discriminate it from an input error.
4. Decide: What evidence would justify changing a withdrawal rule, reservoir policy or monitoring plan? Answer guidance: Compare safety and minimum-service constraints first; then evaluate operating performance, maintenance cost, affected stakeholders and the reversibility of the decision.

Mini field task. Write a one-page technical memo on operation for the seasonal monsoon basin. Include a schematic boundary, a data-quality table, one auditable calculation, a counter-scenario and a clear statement of what the evidence does not establish. Do not present synthetic numbers as observed field data.

11. Flood Control

Core concept. Reservoir attenuation depends on available storage, incoming hydrograph timing, outlet characteristics and downstream synchronization. A simplified routing model with constant outlet capacity captures an introductory conservation effect but cannot model a real spillway rating.

Analysis workflow. Specify maximum safe storage and a daily inflow pulse. Compute downstream release and forced spill each day; check that cumulative inflow minus outflow equals storage change.

Engineering interpretation. Flood-control performance should be assessed over ensembles, not a single hand-picked hydrograph.

Evidence and measurement. Work with an illustrative peri-urban supply district where the available record includes head and storage records. Begin by naming the controlled variable, the response variable and the geographic location of each observation. Check the date, measurement frequency, applicable units and any uncertainty statement. Where a record is missing or derived from a different location, describe the interpolation explicitly rather than silently assuming that it is representative. Before using a numerical method, distinguish a conservation equation from an empirical relation and a legally mandated constraint. Only conservation identities hold without calibration.

From conservation to prediction. A transparent first calculation should reconcile a mass, energy or financial balance as appropriate to the physical mechanism. Convert input quantities into coherent dimensions and display intermediate steps, particularly when mixing depth, area, volume, discharge and time. If a relationship uses a calibrated coefficient, record the period and conditions that support it. Run at least two alternative parameter sets, recording how the predicted outcome changes. A sensitivity experiment identifies leverage but does not by itself assign a probability to uncertainty.

Scenario comparison. Compare a baseline in which current practice continues against an alternative with a different capacity, demand, governance rule or environmental reserve. Where possible, change only one assumption at a time, then carry out a second experiment in which multiple uncertain inputs vary together. Explain what each comparison isolates and what interactions remain hidden. Keep the original measurements, assumptions, units and script version in a reproducible notebook. A repeatable result is necessary but insufficient evidence of a valid model.

Scope and limitations. Do not interpret a screening model as an approved infrastructure design. Confirm whether physical boundary conditions, environmental constraints, public-health requirements and downstream effects appear in the calculation. A conceptual model that omits return flows, aquifer exchange, water quality or operational failures may overestimate usable supply. Sketch a conceptual system with each inflow, outflow, storage and feedback represented. A number without a transparent system boundary is not an auditable engineering result.

Practice questions with answer guidance.

1. Explain: How does the concept of flood control connect to both the quantity and timing of water availability in reservoir systems? Answer guidance: Identify the physically relevant boundary, state the controlling process described above and name one observation that could falsify the assumed behavior.
2. Quantify: Choose two contrasting data periods for the peri-urban supply district, state all units and compute one meaningful normalized comparison; distinguish an absolute difference from a percentage change. Answer guidance: Report the numerator, denominator, observation period, units and rounding precision; do not manufacture a statistical confidence interval without an error model.
3. Diagnose: Suppose a simple model fits the first half of the observation period but fails after a large storm or a prolonged dry season. Which neglected process would you investigate first? Answer guidance: Look for a process whose magnitude or regime changes with the event, then specify a measurement that could discriminate it from an input error.
4. Decide: What evidence would justify changing a withdrawal rule, reservoir policy or monitoring plan? Answer guidance: Compare safety and minimum-service constraints first; then evaluate operating performance, maintenance cost, affected stakeholders and the reversibility of the decision.

Mini field task. Write a one-page technical memo on flood control for the peri-urban supply district. Include a schematic boundary, a data-quality table, one auditable calculation, a counter-scenario and a clear statement of what the evidence does not establish. Do not present synthetic numbers as observed field data.

12. Sedimentation

Core concept. Dams alter sediment continuity. Coarse fractions often deposit near an upstream delta while fine materials may enter deeper water, depending on density currents and residence time. Loss of storage affects both water supply and safety.

Analysis workflow. Convert an annual sediment mass estimate to an approximate deposited volume using dry bulk density and trap efficiency, then update an illustrative capacity estimate.

Engineering interpretation. Bulk density, land-use change, large flood events and sediment management methods make simple linear sediment forecasts uncertain.

Evidence and measurement. Work with an illustrative river-delta corridor where the available record includes population and use records. Begin by naming the controlled variable, the response variable and the geographic location of each observation. Check the date, measurement frequency, applicable units and any uncertainty statement. Where a record is missing or derived from a different location, describe the interpolation explicitly rather than silently assuming that it is representative. Separate a model output from an observation. A discrepancy may reveal a missing process, incorrectly measured inputs, model-structure error or a unit conversion mistake.

From conservation to prediction. A transparent first calculation should reconcile a mass, energy or financial balance as appropriate to the physical mechanism. Convert input quantities into coherent dimensions and display intermediate steps, particularly when mixing depth, area, volume, discharge and time. If a relationship uses a calibrated coefficient, record the period and conditions that support it. Specify an observable metric and the unit in which it would be reported. A useful metric permits comparisons across dates without changing the spatial boundary or its denominator.

Scenario comparison. Compare a baseline in which current practice continues against an alternative with a different capacity, demand, governance rule or environmental reserve. Where possible, change only one assumption at a time, then carry out a second experiment in which multiple uncertain inputs vary together. Explain what each comparison isolates and what interactions remain hidden. Before using a numerical method, distinguish a conservation equation from an empirical relation and a legally mandated constraint. Only conservation identities hold without calibration.

Scope and limitations. Do not interpret a screening model as an approved infrastructure design. Confirm whether physical boundary conditions, environmental constraints, public-health requirements and downstream effects appear in the calculation. A conceptual model that omits return flows, aquifer exchange, water quality or operational failures may overestimate usable supply. Run at least two alternative parameter sets, recording how the predicted outcome changes. A sensitivity experiment identifies leverage but does not by itself assign a probability to uncertainty.

Practice questions with answer guidance.

1. Explain: How does the concept of sedimentation connect to both the quantity and timing of water availability in reservoir systems? Answer guidance: Identify the physically relevant boundary, state the controlling process described above and name one observation that could falsify the assumed behavior.
2. Quantify: Choose two contrasting data periods for the river-delta corridor, state all units and compute one meaningful normalized comparison; distinguish an absolute difference from a percentage change. Answer guidance: Report the numerator, denominator, observation period, units and rounding precision; do not manufacture a statistical confidence interval without an error model.
3. Diagnose: Suppose a simple model fits the first half of the observation period but fails after a large storm or a prolonged dry season. Which neglected process would you investigate first? Answer guidance: Look for a process whose magnitude or regime changes with the event, then specify a measurement that could discriminate it from an input error.
4. Decide: What evidence would justify changing a withdrawal rule, reservoir policy or monitoring plan? Answer guidance: Compare safety and minimum-service constraints first; then evaluate operating performance, maintenance cost, affected stakeholders and the reversibility of the decision.

Mini field task. Write a one-page technical memo on sedimentation for the river-delta corridor. Include a schematic boundary, a data-quality table, one auditable calculation, a counter-scenario and a clear statement of what the evidence does not establish. Do not present synthetic numbers as observed field data.

Unit application and simulation protocol

Laboratory 05 — Reservoir Storage Routing

Purpose and physical meaning. This laboratory isolates one process within water-resources engineering. It is intentionally a limited analytical or bookkeeping model rather than a digital twin of any real basin. The objective is to predict how an output responds to controlled input changes, reproduce the calculation independently and recognize the boundary beyond which the model loses credibility. Use the graph to explore variation, not as evidence of real-world accuracy.

Equation and units. S[t+1] = min(C,max(0,S[t] + I[t]−R[t]−E)); spill tracked. Inputs are base inflow in ML/day, target release in ML/day, and initial storage in ML. Units matter because a factor of 1,000 can arise when litres are confused with cubic metres, or when a daily rate is used directly in a monthly balance. Write out the conversion before comparing scenarios.

Default experiment. Start with the provided values: Base inflow = 25 ML/day, Target release = 30 ML/day, Initial storage = 550 ML. Write a verbal prediction of the first output. Make a hand calculation for the default setting wherever the equation permits it; for a time series, reproduce the first step and check that the total storage change equals the accumulation of fluxes. Only then compare with the displayed model output.

Sensitivity investigation. Move the first slider down to 4 ML/day and up to 65 ML/day, holding the other two at their defaults. Examine the direction and shape of the response, and inspect whether the graph is discontinuous at a physical constraint such as zero storage or an allocation cap. Repeat for the second and third inputs. Export the curve to CSV and estimate a finite-difference sensitivity from adjacent rows; specify the unit of the derivative.

Boundary and validation. Thirty synthetic daily flows use a fixed sinusoidal pulse, a constant target release, evaporation of 1 ML/day and fixed capacity of 850 ML. It is not a real reservoir forecast. A credible validation task would compare the predicted process with a data record measured under comparable conditions and quantify deviations against uncertainties. Do not tune an input to force agreement and then claim independent verification using the same observations. Distinguish a passing arithmetic check from a validated hydrological model.

Challenge. Compare days with spill to days with shortages. Why must mass balance include delivered release, spill and evaporation separately? Report how the answer changes under an extreme but physically plausible input combination, and identify one omitted mechanism that would be important before real implementation. In your conclusion distinguish observations, model assumptions and proposed decisions. If the simulation becomes insensitive to an input, determine whether a maximum/minimum constraint has become active or an equation is coded incorrectly.

Student deliverables. Include a hand calculation, a three-row baseline/low/high scenario table with units, the exported response plot, a one-paragraph discussion of parameter sensitivity, and a one-paragraph critique of model applicability. Name the specific monitoring records and institutional approvals a responsible engineer would seek before designing a project.

Laboratory 06 — Flood-Space Operating Rule

Purpose and physical meaning. This laboratory isolates one process within water-resources engineering. It is intentionally a limited analytical or bookkeeping model rather than a digital twin of any real basin. The objective is to predict how an output responds to controlled input changes, reproduce the calculation independently and recognize the boundary beyond which the model loses credibility. Use the graph to explore variation, not as evidence of real-world accuracy.

Equation and units. S[t+1] = S[t] + inflow − controlled outflow − spill. Inputs are flood pulse peak in ML/day, controlled outlet in ML/day, and starting storage in ML. Units matter because a factor of 1,000 can arise when litres are confused with cubic metres, or when a daily rate is used directly in a monthly balance. Write out the conversion before comparing scenarios.

Default experiment. Start with the provided values: Flood pulse peak = 85 ML/day, Controlled outlet = 45 ML/day, Starting storage = 470 ML. Write a verbal prediction of the first output. Make a hand calculation for the default setting wherever the equation permits it; for a time series, reproduce the first step and check that the total storage change equals the accumulation of fluxes. Only then compare with the displayed model output.

Sensitivity investigation. Move the first slider down to 30 ML/day and up to 150 ML/day, holding the other two at their defaults. Examine the direction and shape of the response, and inspect whether the graph is discontinuous at a physical constraint such as zero storage or an allocation cap. Repeat for the second and third inputs. Export the curve to CSV and estimate a finite-difference sensitivity from adjacent rows; specify the unit of the derivative.

Boundary and validation. Capacity is 800 ML; a synthetic ten-day triangular pulse is followed by lower inflow. Real reservoir outlets are stage dependent and have operating and safety constraints. A credible validation task would compare the predicted process with a data record measured under comparable conditions and quantify deviations against uncertainties. Do not tune an input to force agreement and then claim independent verification using the same observations. Distinguish a passing arithmetic check from a validated hydrological model.

Challenge. How does increasing pre-storm drawdown change the downstream peak and later supply reliability? Report how the answer changes under an extreme but physically plausible input combination, and identify one omitted mechanism that would be important before real implementation. In your conclusion distinguish observations, model assumptions and proposed decisions. If the simulation becomes insensitive to an input, determine whether a maximum/minimum constraint has become active or an equation is coded incorrectly.

Student deliverables. Include a hand calculation, a three-row baseline/low/high scenario table with units, the exported response plot, a one-paragraph discussion of parameter sensitivity, and a one-paragraph critique of model applicability. Name the specific monitoring records and institutional approvals a responsible engineer would seek before designing a project.

Applied project

Project 03 — Reservoir Systems: basin decision memorandum

Decision setting. A synthetic 30-day inflow sequence contains six high-flow days followed by a long dry spell. Operate the same reservoir under a high conservation target and a more aggressive pre-release rule. Compare the peak spill, unmet downstream demand, days below reserve storage and total evaporation. The second rule may reduce flood risk but worsen dry-season supply. Note how the conclusion depends on the quality and lead time of weather forecasts.

Project mandate. Assume a local water commission must make a reversible decision within a limited financial and operational envelope. The project is not to design a real structure; it is to demonstrate reproducible calculations and show when the engineering team must request additional information. Assign a stakeholder team to challenge the baseline scenario and a second team to challenge the preferred intervention. Students should explicitly document disagreement rather than forcing the same preference through all models.

Required tasks.

1. Define the physical system boundary and provide a map or schematic of inflow, outflow, demands and stores.
2. Write an auditable water balance and an explicit list of observations needed to populate it.
3. Compare a baseline, a dry-period stress case and a governance or engineering intervention.
4. Quantify a service-reliability indicator together with one ecological or distributional consequence.
5. State at least two risks not represented by the introductory classroom models.
6. Propose a monitoring plan with observations, frequency, responsible parties and decision triggers.
7. Prepare a transparent summary for people affected upstream and downstream, separating confirmed evidence from assumptions.

Method. Investigate all four syllabus topics of Unit 3: Storage, Operation, Flood Control, Sedimentation. Use the associated pair of simulations as teaching tools, independently verify at least three quantitative results, and do not insert those simulation outputs directly into a permit or construction drawing. When two assumptions point to opposite decisions, describe the range of plausible outcomes and identify which field measurement would most cheaply reduce uncertainty. An optimized number without enforceable operational responsibilities is not a complete solution.

Submission rubric. Technical closure (25%), evidence quality and assumptions (20%), scenario sensitivity and uncertainty (20%), environmental and equity impacts (15%), reproducibility of calculations and charts (10%), and clarity of recommendations (10%). A polished graph cannot compensate for an incorrectly defined basin boundary or missing feasibility constraint.

Unit 03 synthesis and review

As a final check, reconcile the data and assumptions across the four topics rather than treating them as isolated calculations. Prepare a source table showing which information is a measurement, which comes from a model and which expresses a policy decision. Compare what would happen at daily, seasonal and multiannual time scales. Identify one conflict between upstream and downstream objectives and one environmental consequence of the most attractive numerical alternative. Record the monitoring action that would most improve the evidence base. A good water-resources engineer must know what the calculations leave unresolved.

---