I worked on two different projects in the mining industry where accurate modelling of the heat transfer between the ventilation air and chilled water is required. Both the mines, Great Noligwa and Mponeng, are owned by AngloGold Ashanti.
The gold mines in South Africa reach depths of up to 4000m. At this depth the virgin rock temperature rises above 50°C. To allow mine workers to work at the active mining location the ventilation air must by cooled to 25°C.
The mining industry is continuously looking into more efficient systems to cool the mining environment in order to save costs. Currently most mines use bulk air coolers to cool the air going down the shaft and cooling cars to cool the ventilation air in the haulages and cross-cuts.
For both of the simulation projects we needed a ventilation air and chilled water network. The networks are integrated with each other through the heat exchange at the bulk air coolers and cooling cars.
My work was to create a cooling car model to accurately simulate the heat exchange between the ventilation air and the chilled water. The manufacturer's performance data and the geometry of the cooling car were known. The manufacturer's performance data included the heat transfer values for different water and air flow rates at three different water inlet temperatures. The cooling car had a finned tube heat exchanger arrangement and the valuable geometrical input data that were available was the number of passes, number of tubes, length of the tube passes and the inner and outer diameters of the tubes.
Example of a cooling car as manufactured by Manos Engineering
The ability to do accurate modelling of the psychrometric behaviour of the humidity in air was very important to obtain accurate results. Flownex (a system based thermal fluid simulation software package) has the ability to accurately model a two-phase fluid and to also combine it with another fluid to create a mixed fluid. In the simulation the fluid as used was created by combining water (two-phase fluid) with air (an incondensable fluid). Flownex also had a finned-tube heat exchanger component which was used to model the cooling car.
The remaining input data required was the heat transfer characteristic chart of the heat exchanger. This chart is a plot of the Colburn j factor versus the Reynolds number and has a negative exponential relationship (y=x^-b). The Colburn j factor is used to determine the heat transfer coefficient over the fins. The chart was successfully reverse-engineered by matching the manufacturer's performance data with the simulated results for the air flow rate going through the design point. The matching was done by adjusting the coefficients in the exponential relationship until the results were acceptable.
After determining the heat transfer characteristic chart, the cooling car component was used to predict the heat transfer for different air flow rates. Comparison to the manufacturer's data showed that all the performance values predicted were easily within a 5% error margin. The pressure drop characteristics for the tube side and fin side were also easily matched.
The result was a simulated cooling car component that was verified and validated with the manufacturer's data. This component was successfully used in an integrated systems network. This network was used to identify the possible energy saving opportunities and the integrated effect thereof. The work done at Mponeng resulted in savings of 2.5MW and received a "Special Award" in the Innovation category at ESKOM's Eta Awards.