ENFOR - Heat Load Prediction and Energy Systems Optimization

Heat Load Prediction and
Energy Systems Optimization (PRESS)

The PRESS (Prognosis and Energy Control System) is used for load forecasting, temperature control, and optimization of district heating systems. PRESS can be used in many different ways depending upon your main focus. Some installations use PRESS for on-line forecasts of the heat load, whereas other installations are using PRESS for an automatic control of the supply temperature. Also a total optimization of the heat production can be embedded in the PRESS system.

History

The development of PRESS started in 1988. The original idea, as proposed by some district heating systems, was to used the data already available in most district heating systems for on-line forecasting, temperature control and optimization. In many years the system was mostly developed for supporting research on heat load forecasting and optimal temperature control, but since late 90'ties the system has been used operationally by a increasing number of district heating systems.

PRESS

Characteristics

PRESS is built on artificial intelligence, and hence the system automatically calibrates to the actual situation. For instance the models for forecasting the heat load automatically takes into account changes in the number of customers connected, or changes in the user profiles, and the models for automatic control adapts automatically to changes in the network system. Hence there is no need for a very time demanding detailed input of the network.

PRESS takes a number of input data, like the observed heat consumption, return temperature, ambient air temperature, wind speed, meteorological forecasts, etc., and, as an example, PRESS provides forecasts of the heat load up to 120 hours ahead, and determines the optimal supply temperature in order to ensure the most optimal situation, both with respect to a minimization of the energy losses and, at the same time the system ensures a sufficiently high supply temperature at all the consumers.

For further technical details about PRESS and including some examples of the benefits of using the system we refer to the section for technical papers or the following PDF-presentation.

Input to PRESS

Depending on the configuration of PRESS the following data is taking into account:

Online measurements of heat load
Supply and return temperature
Water flow at the production plants
Local measurements of ambient air temperature, wind speed, solar radiation, etc.
Meteorological forecasts of the local air temperature, wind speed, etc.

Besides these data PRESS automatically takes into account the systematic behavior of the consumers heat consumption. Also the heat stored in the distribution network is taken into account.

Modules in PRESS

The total PRESS system consists of a number of modules:

Data collection and validation
On-line heat load forecasting
On-line network temperature control
On-line statistics summarizing the operation of the district heating systems

Optimization – tailored to the need of the individual

PRESS can be used in many different ways, depending upon your main focus. Some installations use PRESS for heat load forecasting only, whereas other installations are using PRESS for an automatic control and optimization of the supply temperature.

Module: Data collection and validation

Since PRESS is an on-line system, the data collection and validation module is needed in all cases. The module collects the necessary data, detects possible errors in the data, and validate the data for the on-line application in the subsequent modules.

Module: On-line heat load forecasting

The heat load forecasting is based on meteorological forecasts of primarily ambient air temperature, but the system is also able to use forecasts of radiation and wind speed, if available. Typically, but depending on configuration, the system forecasts the hourly heat load up to 120 hours ahead.

In cases where meteorological forecasts are unavailable for some reason, the PRESS system uses local measurements of ambient air temperature, wind speed, etc. instead. In this case the maximum prediction horizon is, say, 24 hours.

One of the aspects complicating heat load forecasting is the fact that the heat dynamics of the buildings seriously affect the heat demand on an hourly basis. The PRESS system automatically takes this smoothing effect into account. Furthermore, the system continuously adapts to the actual situation by continuously monitoring the consumption. In this way the systems automatically adapts to things such as

Changes in consumer behavior
Changes in the number of consumers
Changes in the meteorological models
Changes in the thermal characteristics of the buildings
Changes in the network

Module: On-line network temperature control/optimization

To reduce heat loss in the distribution network the temperature in the district heating network should be kept low while at the same time keeping the temperature sufficiently high at the consumers and respecting pumping costs.

The online network temperature optimization module achieves this goal by monitoring the temperature at a number of critical points in the network and respecting the pumping costs using the heat load forecasts.

The controller used is very advanced in several aspects, e.g. it is able to handle the time-varying time delay from supply point to the critical points and it continuously adapts to system changes.

Module: On-line statistics

Using consumption and temperature data this module continuously gathers information about the characteristics of the system and presents this in a comprehensive form. This information includes:

Duration curve of the consumption
The energy signature, i.e. the expected consumption as a function of air temperature.
The diurnal variation of the consumption.

In this way the persons operating the system can easily monitor the system and take action if undesirable changes occur.

Module: Optimization

This module is typically tailored to the individual district heating system. However, very often optimization related to the used of heat storage tanks is relevant. Other problems often seen are optimized scheduling of the heat production.