In this paper, we present findings from a study of the vertical and horizontal temperature distribution in a water tank to simulate a bulk petroleum tank. The temperatures were measured using discrete integrated circuit temperature sensors in a multiplexed sensor network. The experimental study found that there was a variation in the horizontal and vertical temperature profiles of up to 0.7 °C and 4.3 °C respectively. The horizontal temperature profile can be attributed to heating of the equatorfacing surface from solar radiation. The temperature profile presents an uncertainty in the measurement of tank content volume and can cost tank owners many thousands of dollars.
IntroductionCorrect inventory control of liquid contents in bulk storage containers require accurate measurements of the liquid's volume. Liquid volume is determined by measuring the liquid height and its volume is calculated by geometric means. Accurate calculations of volume require an accurate determination of temperature. As the content's temperature fluctuates, there is a corresponding fluctuation in it's density, hence a change in volume. This uncertainty in volume can cost tanker owners many tens of thousands of dollars. A one degree Celsius change in temperature can cause a 0.1% volumetric change in petroleum. In a 10,000 m 3 tank, this represents an uncertainty of 10 m 3 or 10,000 L. At current oil prices of USD$120/barrel this equates to an uncertainty cost of USD$7,560. Over a period of time, this uncertainty cost can be quite substantial. Multiply this over multiple tanks in a typical tank farm and it is clear that there is a need for accurate temperature monitoring by bulk tank owners. This work outlines the development of a volumetric temperature sensor network to measure both the vertical and horizontal temperature gradients to study the effects of these gradients on the average tank content temperature.