In commercial nuclear power plants spent fuel assemblies are usually stored in actively cooled water pools. The continuous decay heat release represents a potential risk in case of a station black out scenario. Thus two-phase passive heat removal systems are a key technology to enhance the safety of nuclear power plants. Such systems work only by the energy provided from the heat source, e.g. by the maintenance of a natural convection cooling. A heat transfer loop using air as an unlimited heat sink consists of a primary heat exchanger in the spent fuel pool water and a secondary heat exchanger located in ambient air. Thus the measurement of the heat flux, which gets transferred from the pool to the ambient air, is an important task. If one would measure heat flux, flow rates and temperatures in many positions by help of local probes, the natural flow would get strongly disturbed. For that reason we introduce a heat flux measurement around the secondary heat exchanger located in ambient air, which applies temperature and velocity measurement by an anemometric principle.A 6.5m long flow channel with an electrical heated finned tube heat exchanger was set up at the TOPFLOW facility at HZDR. Since the tubes of a heat exchanger would be tilted in a passive heat removal system, i.e. to allow drainage of the condensed heat transfer medium, different tiled angles were adjusted to 0° (horizontal), 20°,30° and 40°. The frontal velocity was varied between 0.5ms and 4ms and three thermocouples were placed up- and downstream of the heat exchanger respectively. A Temperature Anemometry Grind Sensor (TAGS) was located downstream the heat exchanger. It consists of a wire grid with platinum resistance elements, which are placed in the small sub-channels of a flow straightener to generate laminar flow profiles. Two methods were used to calculate the heat flux: arithmetical average and weighting of the flow area. The results of velocity was compared with the average velocity measured by the volume flow control and out of the velocity and temperature the heat flux was calculated and compared with electrical supplied heat flux. The calculated average velocity measured by the TAGS corresponds well with the velocity measured by the volume flow controller up to approximately 3ms with a maximum deviation of ±5%, but underestimates the velocity measured by the volume flow controller at higher velocities. The heat flux was calculated by five methods, 1.) from the three thermocouples up- and downstream of the heat exchanger, 2.) from the average temperatures measured by the TAGS, 3.) from the weighted temperature measured by the TAGS, 4.) from the average temperature and velocity measured by the TAGS and 5.) from the weighted temperature and velocity measured by the TAGS. In this order the accuracy of methods increases compared to the electrical supplied heat flux. For the last method the maximum deviation was 6.5% for all tilt angles. This measurement concept determines the heat flux without disturbing the flow in the loop.