P25
By Peter Friis Østergaard, Senior Specialist, and Jakob Fester, Consultant, Danish Technological Institute
The integration of remotely read district heating meters on an hourly or a daily basis has opened up new possibilities for data-driven optimization of the entire distribution network, from fault detection to heat production. The promise of being able to lower costs based on optimized heat production and fast leak detection has enabled a roll-out of smart meters to all corners of many district heating networks. However, if these meters are only used for annual settlements, an old-fashioned, manually read meter would do almost as good, without the added cost of adding a transmitter and building up a data network. To ensure a fruitful investment, the added value of the online data must be brought into play. At Danish Technological Institute we are actively working towards using the meter data, as well as data from additional sensors, to add value for the district heating supplier and the customer. The first steps focus on customer classification and fast fault detection combined with additional information by low-cost high-frequency pressure and temperature sensors that can easily be installed and moved between strategic locations in the grid. This work is currently conducted through two individual Danish and European projects and a performance contract with the Danish Agency for Institutions and Educational Grants. Expectedly, the tools developed will save a significant amount of manhours spent today on manual data inspection at small and medium-sized district heating companies, as well as leading to savings relating to faster repair and readjustments of incorrect settings of installations. BUILT FOR ACCOUNTING One of the challenges of using smart meters for optimization of district heating is the fact that smart meters are still constructed mainly as a meter intended for settlements of accounts. They are designed to provide readings used for the distribution networks to send bills, not as data loggers suited for acquisition of high-precision data. To exemplify this, the temperature sensors in the meter are all calibrated by the manufacturer, matched with a similar sensor and installed together in one meter. As such, it should be possible to detect the calibrated temperature at the installation, along with uncertainties for the measurement.
Unfortunately, the calibration of the sensor is most often used only during the production at the factory in order to comply with current documentary standards. As soon as the meter is installed at the consumer, the information from the calibration is disregarded. The consequence is that the accuracy of the direct temperature reading on the meter is strongly compromised. Even if the calibration parameters were available, the data transfer from the meter to the database is often problematic. Transferred data is often rounded off to the nearest integer, and sometimes consists of derived values, meaning that the data resolution can be much lower than what would have been possible.
Finally, data points are often missing which creates challenges when complete data sets are needed for the analysis.
LACK OF CONSENSUS A widespread consensus on data formats, data ownership, transfer protocols, etc. may be lacking for smart meters. Individual manufacturers often promote their own special abilities during tenders, and as a district heating distributor it may be difficult to figure out, what is relevant and what is not. The district heating distributors want value for money, but since it is hard to see where the value might be found, the choice of meter manufacturer is difficult. This situation is not all bad since having multiple manufacturers competing on providing the best product also has the consequence that innovation moves faster, resulting in better solutions in the long run. To assist the small andmedium-sized district heating suppliers in gaining the benefits from the smart meters, without necessarily being locked to a single manufacturer, Danish Technological Institute is developing a tool, which can be used for classifying the customers, making heat load predictions and performing a general check of the district heating network, independent from the manufacturer of the meters.
www.dbdh.dk
J O U R N A L N 0 . 3 / 2 0 1 9
Made with FlippingBook - Online Brochure Maker