Converting legacy network data into a simulation-ready district heating model is one of the most underestimated challenges in heat network hydraulic modeling. On the surface, the task sounds straightforward: gather existing records, import them into the software, and begin running scenarios. In practice, the data landscape that most district heating utilities inherit is fragmented, inconsistently documented, and often decades old. The gap between what exists in records and what a physics-based simulation model actually needs is where projects stall — and where the quality of the eventual model is decided, long before the first simulation runs.

This article walks through the key considerations that determine whether a data conversion effort produces a genuinely useful thermal network simulation tool or a model that looks complete but behaves unreliably under real operating conditions. For utilities planning network expansions, evaluating production mixes, or building toward a digital twin, understanding these fundamentals is the necessary foundation.

What legacy data quality means for simulation accuracy

A district heating simulation model is only as accurate as the data it is built from. Physics-based simulation engines calculate flow rates, pressure gradients, and temperature distributions based on the physical properties of the network — pipe diameters, lengths, roughness coefficients, substation heat demands, and pump characteristics. When any of these inputs are approximate or missing, the model compensates with assumptions, and those assumptions compound across the network.

The practical consequence is that a model built on poor legacy data may produce results that look plausible during normal operating conditions but diverge significantly from real behavior during peak demand, low-load periods, or fault scenarios. For a utility using district heating software to evaluate a network expansion or test a new pumping strategy, that divergence is not a minor inconvenience — it is a decision risk. The value of simulation lies in its ability to replace real-world experimentation with safe, repeatable analysis. That value disappears when the model cannot be trusted.

The hidden complexity of district heating network data

District heating networks accumulate data across multiple systems and timescales. Asset registers, GIS databases, maintenance records, SCADA logs, and engineering drawings may all hold relevant information — but rarely in a form that maps directly to the data structure a simulation model requires. Pipe records may specify nominal diameter rather than internal diameter. Substation records may document installed capacity without reflecting actual consumption patterns. Network extensions built in different decades may follow entirely different documentation conventions.

There is also the question of data provenance. In older networks, physical surveys and as-built drawings may be the primary source of truth, and those documents may not have been updated to reflect subsequent modifications. Pipe replacements, substation upgrades, and valve additions carried out over years of operation create a version of the network that exists in reality but not in any single record. Identifying and resolving these discrepancies before model construction begins is time-consuming work — but it is far less costly than discovering them after calibration has started.

Key data types every simulation model depends on

Building a simulation-ready model for a district heating network requires a specific set of data inputs. Not all of them are equally available in legacy records, and understanding which are critical versus which can be estimated guides the data collection effort effectively.

• Pipe geometry and material properties: Internal diameter, length, and roughness coefficient for every pipe segment. These determine friction losses and flow resistance across the network.
• Substation demand data: Heat consumption profiles for connected buildings, ideally at hourly resolution across a full annual cycle. Peak demand values alone are insufficient for dynamic simulation.
• Pump and valve characteristics: Pump curves, control logic, and valve positions are essential for accurate pressure and flow modeling, particularly in networks with multiple production sources or pressure zones.
• Supply temperature profiles: The temperature of hot water leaving the production plant varies by season and operating strategy. Accurate supply temperature data is required for thermal simulation and heat loss calculations.
• Network topology: The connectivity of the pipe network, including all branches, loops, and dead ends. Topology errors — missing connections, incorrect node assignments — are among the most common sources of simulation error.

In practice, pipe geometry and network topology are usually the most complete elements of legacy data, while substation demand profiles and pump characteristics are frequently the least well documented. Prioritizing data collection efforts around these gaps early in the project saves significant time during calibration.

Common pitfalls when converting network data to a model

Several recurring problems appear when district heating network data is converted into a simulation model without sufficient preparation. Recognizing them in advance reduces the risk of building a model that requires extensive rework.

Incomplete topology: GIS data exported for modeling often contains connectivity gaps where pipes that appear visually connected are not logically linked in the data. These gaps produce isolated network segments that the simulation engine cannot solve. A systematic topology check before import is essential, not optional.

Inconsistent unit systems: Legacy records from different eras or different departments may use different unit conventions — millimeters versus inches for pipe diameters, kilowatts versus megajoules for demand values. Importing data without normalizing units produces errors that can be difficult to trace once the model is assembled.

Static demand values treated as dynamic inputs: A common shortcut is to assign each substation a single peak demand value and treat it as a fixed load. This approach produces a model that may calibrate acceptably under peak conditions but fails to represent the network’s behavior across the operating range. District heating optimization software requires demand profiles that vary with time and outdoor temperature to produce meaningful scenario results.

Unchecked pipe age and roughness assumptions: Older pipes accumulate internal scaling and corrosion that increases hydraulic roughness over time. Using manufacturer roughness values for pipes that have been in service for decades introduces systematic error in friction loss calculations across the network.

A structured approach to building a simulation-ready model

The most effective approach to data conversion follows a defined sequence that separates data preparation from model construction. Attempting both simultaneously typically results in a model that is difficult to calibrate because it is unclear whether discrepancies originate from data quality or model configuration.

The first phase is a data audit and gap analysis. This means systematically reviewing all available data sources against the input requirements of the simulation model, identifying what is complete, what is approximate, and what is missing entirely. The output is a prioritized list of data gaps and a strategy for addressing each one — through targeted surveys, engineering estimates with documented assumptions, or data derivation from related sources such as billing records or SCADA logs.

The second phase is data preparation and normalization. All source data is converted to consistent units, checked for topological integrity, and structured to match the import format of the district heating network design software. This phase often reveals additional gaps that were not apparent in the audit, particularly in network topology where visual inspection of GIS data can be misleading.

The third phase is staged model construction. Rather than building the full network in a single pass, an effective approach starts with the main transmission network — the primary supply and return mains connecting production to major distribution nodes — and validates this backbone before adding secondary distribution and individual substations. This staged approach makes it significantly easier to isolate and correct errors before they propagate through the full model.

Model calibration follows, comparing simulation outputs against measured operating data from SCADA or field surveys. Calibration is an iterative process, and the quality of the legacy data directly determines how many iterations are required before the model meets an acceptable accuracy threshold. For utilities working through this process for the first time, Fluidit’s expert consulting services can accelerate the calibration phase by applying a structured methodology developed across many network conversion projects — bridging the gap between raw data and a model that genuinely reflects real network behavior.

From static model to operational digital twin

A simulation-ready model built through a rigorous data conversion process is a valuable planning asset. It supports network expansion analysis, production strategy evaluation, and supply security assessment. But its potential does not end there. The same model, connected to live operational data, becomes a digital twin — a continuously updated representation of the network that supports day-to-day operational decisions alongside long-term planning.

The transition from a static district heating model to an operational digital twin depends on the quality of the underlying model. A model built on incomplete or poorly calibrated data cannot be extended into real-time operation reliably. This is one of the strongest arguments for investing in data quality and rigorous model construction from the outset, even when the immediate objective is a static planning model. The work done to build a trustworthy simulation model is the same work that enables the digital twin.

Fluidit Heat is purpose-built for this progression — from initial network data import through to physics-based scenario simulation and, where the data infrastructure supports it, real-time model integration. The platform’s architecture is designed to grow with the utility’s data maturity, so that the model built today can evolve as operational data becomes richer and more connected.

For district heating utilities managing aging networks, planning for new heat sources, or working toward tighter emission targets, a well-constructed simulation model is not a technical exercise for its own sake. It is the analytical foundation on which better decisions are made — decisions about where to invest, how to operate, and how to demonstrate the system’s performance to regulators, customers, and stakeholders. Getting the data right from the start is what makes that foundation solid.