Energy efficiency
Building energy management

Posted on Thursday, July 14, 2022 by Mads Wraa Hyttel

How to improve data quality in a BMS system

Accuracy of temperature and flow rate, temperature offset and valve control form a perfect recipe to manage low delta T syndrome. Now you can get it all packed in a smart meter solution, It’s time to know!

Building management systems (BMS) are widely used in buildings to control and monitor electrical and mechanical equipment responsible for a long range of services in the building. This typically includes the cooling system in the building. If cooling is supplied using a chilled water system a BMS system will use the temperature and flow measurements from the installed cooling meters to obtain the expected comfort in the building at the lowest possible energy consumption. This is especially the case when trying to solve the problem known as “low delta T syndrome” where the cooling system is in an undesirable state where there is a high flow and low temperature difference. Accurate data on temperature and flow rate from the cooling meters are therefore essential for efficient operation of the BMS system.

In this blog we will therefore discuss what can be done to improve the data quality in a BMS system when it comes to temperature and flow measurements. The starting point is that the 2-wire Pt500 sensors that are typically used on cooling meters are not completely accurate. For instance, the sensor element inside the Pt500 sensor has an accuracy of ± 0.3 K when used for a temperature range of 0…20 °C. Furthermore, the resistance of the cable of a 2-wire sensors increases the measured inlet or outlet temperature with up to + 0.8 K if the cable length is 10 m. In total, the measured inlet and outlet temperatures in BMS systems are typically higher than the actual temperature and can be up to 1 K higher than the actual temperature. Since a chilled water system is designed to operate at very specific inlet and outlet temperatures this offset can lead to higher than necessary energy consumption of the cooling system.

Figure 1: Accuracy when measuring inlet or outlet temperature on a cooling meter.

The latest generation of Kamstrup cooling meters comes with a novel feature called “temperature offset”. As part of the European standard EN1434 for thermal energy meters each temperature sensor is measured at various temperatures before the pairing with a second temperature sensor is performed. The offset explained in the above figure is therefore easy to determine based on these measurements and is subsequently added to the cooling meter. The result is a cooling meter where the inlet and outlet temperature is measured with an accuracy of ± 0.1 K. With the new Kamstrup TemperatureSensor 63 the calculated temperature offset is also written on the label of the temperature sensor making the offset value transparent for everyone and easy to insert in the meter if some reason the temperature sensors must be changed.

Figure 2: The temperature offset function in Kamstrup cooling meters uses data from previous measurements of the temperature sensors to obtain a high accuracy of ± 0.1 K when measuring inlet or outlet temperatures.


The benefits of measuring the inlet and outlet temperature with an accuracy of ± 0.1 K are several. First of all, it solves the general problem explained above that the measured inlet and outlet temperatures has a tendency to be higher than their actual temperature. Furthermore, it is now possible to compare all temperature measurements inside the BMS system since any large offset have been eliminated. This is especially a problem if there is a variation of temperature sensor cable lengths inside the BMS system. For instance if one cooling meter is with 10 m temperature sensor cables and other cooling meters with 3 m cables the difference between two cooling meters can easily be 0.5 K as shown in Figure 3.

Figure 3: A typical scenario when using standard 2-wire temperature sensors for measurement of inlet temperatures.


Sometimes it is decided to use only 4-wire sensors to improve accuracy of the temperature measurements. This also eliminates the increase in temperature caused by the cable length but does not remove the accuracy of the sensor element of ± 0.3 K. And 4-wire sensors are both more expensive to buy and to install.

A Kamstrup cooling meter with temperature offset makes it possible to achieve an even better accuracy with a 2-wire sensor even at a lower cost eliminating the need for 4-wire sensors if a maximum cable length of 10 m is sufficient. If longer cable lengths are necessary the 4-wire Kamstrup TemperatureSensor 83 also supports the temperature offset feature ensuring that the accuracy of ± 0.1 K on all cooling meters. The difference between using standard 4-wire sensors and 2-wire sensors with offset adjustment is shown on Figure 4.

Figure 4a: Deviation in measured inlet temperatures when using standard 4-wire sensors

Figure 4b: Deviation in measured inlet temperatures when using standard 2-wire sensors and temperature offset adjustment in the cooling meter.


The “low delta T syndrome” where there is a high flow and low temperature difference in the chilled water system is unfortunately very common. A good starting point to solve this problem is to use temperature sensor pairs that measures the temperature difference, delta T, with high accuracy. The Kamstrup temperature sensors are paired using multiple measurements as part of a high volume production ensuring a typical accuracy that is better than the accuracy required in the European standard EN1434 for thermal energy meters. One solution to solve the low delta T syndrome is use a valve that is managed by a PQT controller which then opens and regulate the valve based on specific setpoints for e.g. power, flow or temperature difference using actual measurements made by the cooling meter as explained in this blog. The efficiency of this solution therefore again depends on the accuracy of used temperature and flow measurements. This further emphasizes the need for temperature measurements with the high accuracy obtained using the temperature offset feature. A standard class 2 flow sensor has an accuracy of 2-4% depending on the flow rate. Kamstrup also offers a special flow sensor with an accuracy of 0.5 % at nominal flow qp. This increased accuracy of the measured flow will improve the efficiency of the PQT controller since several functions inside the PQT controller depends on the measured flow.

To sum up

A BMS systems rely on accurate measurements of both flow, inlet and outlet temperatures to function properly and obtain the expected energy consumption of a chilled water system. This is especially the case if a valve has been installed to solve the low delta T syndrome. Unfortunately, the standard temperature sensors used on most cooling meters have typically only been optimized for measuring the temperature difference since this is used for billing. To solve this problem Kamstrup has introduced a unique solution that significantly improves the accuracy of the measured inlet and outlet temperatures in a very cost effective way. A high accuracy of the measured flow is ensured by our market leading ULTRAFLOW® technology. A Kamstrup cooling meter is therefore the right choice if you want to ensure an improved data quality in your BMS system.


Related blog posts