🎓 Lesson 11
D5
Modeling Realistic Demand Response to Pressure Changes
When water pressure changes in a pipe, how much water users actually take—like from faucets or irrigation—changes too, and engineers must predict that change to keep the system working well.
🎯 Learning Objectives
- ✓ Calculate pressure-dependent demand at a node using the exponent-based PDD model
- ✓ Analyze network performance under reduced pressure scenarios by comparing PDD vs. fixed-demand simulations
- ✓ Design pressure management zones by applying PDD constraints to meet minimum service pressure requirements
- ✓ Explain how leakage flow rates scale with pressure using the orifice-based leakage model
📖 Why This Matters
In real-world water systems—especially aging infrastructure or hilly terrain—pressure fluctuates constantly. During peak demand or pump failures, pressure can drop below 20 m, causing taps to sputter and fire flows to fall dangerously low. Ignoring pressure-dependent demand leads to overly optimistic hydraulic models: you might think the system meets demand when, in fact, customers receive only 40% of their requested flow. This lesson bridges theory and practice—ensuring your models reflect physics, not assumptions.
📘 Core Principles
Demand in water networks isn’t static—it’s governed by fluid mechanics at the point of use. At a tap or sprinkler, flow is driven by the pressure head difference between the pipe and atmosphere. Below the required minimum operating pressure (typically 10–15 m), fixtures cannot deliver full rated flow; above it, flow plateaus. Leakage behaves similarly but follows an orifice equation: flow scales with the square root of pressure difference across pipe wall defects. Two complementary models dominate practice: (1) the exponent-based PDD model (used for customer demand), and (2) the pressure exponent leakage model (used for background losses). Both rely on empirically calibrated exponents—usually between 0.5 and 1.5—reflecting fixture type, valve geometry, and leak orifice shape.
📐 Key Calculation
The exponent-based pressure-dependent demand model expresses actual demand as a function of available pressure relative to a reference pressure. It’s widely implemented in EPANET and industry-standard simulation tools.
💡 Worked Example
Problem: A residential node has a base demand of 3.2 L/s at reference pressure P_ref = 25 m. The pressure exponent α = 0.75. During a simulation, the computed nodal pressure is 14.5 m. Calculate actual demand.
1.
Step 1: Verify pressure is above minimum (P_min = 10 m) and below P_ref — yes, 10 < 14.5 < 25.
2.
Step 2: Apply PDD formula: Q_actual = Q_base × (P_node / P_ref)^α = 3.2 × (14.5 / 25)^0.75.
3.
Step 3: Compute ratio: 14.5/25 = 0.58 → 0.58^0.75 ≈ 0.652 → 3.2 × 0.652 = 2.086 L/s.
Answer:
The actual demand is 2.09 L/s, which is ~65% of base demand—consistent with typical residential response at 58% of reference pressure.
🏗️ Real-World Application
In the 2022 pressure management pilot in Cape Town’s Tygerberg zone, engineers used PDD modeling to redesign PRV (pressure reducing valve) setpoints. Fixed-demand models predicted 99.2% service coverage; PDD simulations revealed that 18% of households received <15 m pressure—and thus only ~55% of intended flow—during evening peaks. By lowering PRV setpoints from 55 m to 42 m and adding booster stations in low-elevation pockets, they achieved uniform 35–45 m residual pressure, cutting non-revenue water by 11% and improving equitable supply without increasing pumping energy.
🔧 Interactive Calculator
🔧 Open Water Distribution Network Analysis Calculator📋 Case Connection
📋 Calibration of Lagos Metropolitan Water Network
Persistent model–field mismatch (>25% pressure error) due to undocumented pipe replacements and unaccounted demand growt...
📋 Real-Time Pump Scheduling for Barcelona’s Tertiary Reservoir System
Excessive energy costs (34% of OPEX) and frequent low-pressure complaints during peak afternoon hours
📋 Water Quality Model Validation for Singapore’s Deep Tunnel Sewerage System (DTSS) Supply Branch
Disinfectant residual dropping below 0.2 mg/L at farthest nodes despite design dosing; suspected wall reaction dominance