District Cooling Network Optimization

Two-Stage Stochastic Programming for District Cooling Network Optimization Under Uncertainty

Authors: Manfredi Neri, Elisa Guelpa, Vittorio Verda
Source: ScienceDirect – Applied Thermal Engineering

Introduction

The optimization of district cooling networks is crucial due to the high capital costs associated with their installation and operation. Traditional deterministic models often fail to account for uncertainties in cooling demand growth and electricity prices, leading to inefficiencies in system design.

A recent study by Manfredi Neri, Elisa Guelpa, and Vittorio Verda introduces a two-stage stochastic programming model that optimizes the design of district cooling systems (DCS) under uncertain demand and cost fluctuations. This novel approach provides a more cost-effective and flexible solution compared to conventional deterministic models.

---

Key Findings

1. Cost Savings & Payback Time:
   • The stochastic model reduces costs by up to 5% compared to deterministic methods.
   • The expected payback time is three years shorter, making investments in district cooling systems more attractive.
   • The Net Present Value (NPV) is up to 54% higher, highlighting the economic benefits of a flexible approach.
2. Optimal System Expansion Strategy:
   • Instead of oversizing the network at the start, the model suggests installing a smaller system initially and expanding it later based on future conditions.
   • This approach minimizes financial risk if electricity prices or cooling demand do not increase as anticipated.
3. Impact of Key Uncertainties:
   • The study found that electricity costs and cooling demand growth have the most significant impact on network design.
   • Higher electricity costs encourage a wider adoption of district cooling systems, while lower costs lead to a preference for individual chillers.
4. Flexible vs. Rigid Model Comparison:
   • A flexible model allows for additional pipelines in future expansions, leading to lower costs and better scalability.
   • The rigid model (which restricts adding new pipes in the same branches) results in higher initial investments and lower adaptability to future changes.

---

Why This Study Matters

With global cooling demand continuing to rise, cities need efficient and cost-effective solutions for cooling infrastructure. District cooling networks can reduce energy consumption by up to 40% and cut lifecycle costs by 20%. This study provides a data-driven optimization framework to help urban planners and decision-makers design sustainable and adaptable district cooling systems.

By incorporating uncertainty in decision-making, this approach significantly reduces financial risk, making district cooling a more viable and attractive investment for cities and businesses alike.

---

Read the Full Research

For a detailed analysis, methodology, and case study results, read the full paper on ScienceDirect:
👉 Two-Stage Stochastic Programming for the Design Optimization of District Cooling Networks Under Demand and Cost Uncertainty

This study is open access under a Creative Commons license, allowing free access to all readers.