One of the most critical requirements of a Ventsim™ model is that it accurately represents the ventilation conditions in your mine. Without an accurate representation of current conditions, future design using the model will not be accurate and could provide misleading or incorrect results.
For this reason, validating your model with current actual ventilation data and conditions is very important.
How accurate should a Ventsim™ model be? There is no correct answer, but every engineer and consultant should have a target to aim for based on what the model is required for. A model that is to be used to predict future expensive infrastructure and fans should target a higher accuracy than perhaps a long term life of mine draft.
Accuracy represents the difference between modelled airflows and pressures, and the actual measured airflows and pressures. Typically most consultants will aim for around 95% accuracy in the primary airways, but this is highly dependent on the quality of data that is put into the Ventsim™ model and may be difficult to obtain unless a complete pressure survey is done to establish accurate resistances. For less critical applications, less the 90% may be acceptable, particularly in lower airflow areas.
In theory, if perfect input data is used within a Ventsim™ model, the output simulation should be also perfect. The Hardy Cross algorithm used by Ventsim™ ensures accuracy can be achieved (within the tolerances set for the simulation solution error). Unfortunately the airway sizes, friction factors, shock losses and fan performances are often not exactly known and can be difficult or time consuming to measure in some areas. For that reason, engineers will often use standard design sizes and default friction factors and shock losses to design a model. While not as good as measuring actual resistances with pressure surveys, this type of data will commonly still achieve 85 – 90 % accuracy, particularly if airway sizes can be more accurately validated with survey data that can help identify variation from standard design sizes.
Some tips on creating accurate models:
- Where airway sizes or resistance cannot be accurately measured, use the mine design to help create the Ventsim™ model, and use the actual survey data to improve the Ventsim™ model airway size estimation. Most mines will commonly overbreak the design size so that final airway size may be 10 – 15% larger. If this is not accommodated in the Ventsim™ model, the model may over predict the pressure or under predict the airflow.
- Don’t forget to consider Shock Losses, particularly at major intersection and junctions with lots of airflow. Shock Losses can commonly add 10 – 15% additional pressure requirements to a fan and must be considered. Funnily enough, engineers who do not consider over-break from design sizes, also forget to consider shock losses, and the effect somewhat cancel each other out across the model – but not always in the correct places.
- If you cannot perform a pressure / resistance survey across the mine, then at least try to measure some example resistances in main airways. These can be used to derive friction factors which can then be applied to other similar airways, and are probably better than using standard default friction factors.
- Ensure all simulation parameters and settings are accurate. This step requires a systematic audit or all information used and will be covered in more detail in a future article. Each major ventilation assumption (resistance, friction factor, airway size, fan curve, shock losses, simulation settings such as compressible flows and surface temperatures and barometric pressures) needs to be reviewed for accuracy. Most Ventsim™ data can be displayed in different colours so it is easy to examine an entire model for colours that are not consistent with what you would expect.
- If the mine is known to have a strong natural ventilation pressure presence, then this will need to be included in the model. While the Ventsim™ automatic natural ventilation option can be used, unless a very accurate heat simulation model is made of the mine this will not give the correct result. In this case, it may be best to TURN OFF automatic natural ventilation, and use FIXED PRESSURES in the surface airways to simulate known natural ventilation pressures.
Validating the Simulation versus Survey Data
The best way to validate the Ventsim™ simulation is to include the actual measured results of ventilation surveys as ‘Notes’ within the model. The notes can then be compared side-by-side with the simulated results, showing any discrepancies. An even easier solution is to use VentLog™ software offered by Chasm Consulting that allows this to be done more easily. Ventilation survey data can be entered into the VentLog™ software, and automatically overlayed on the Ventsim™ model.
In addition, mine survey topography can be imported as a reference and overlaid on the Ventsim™ model design. This will show discrepancies in design sizes and airway locations that may warrant further investigation.
Don’t always expect the Ventsim™ model to exactly match the underground ventilation survey data. Often the survey data itself may have errors in measurement – it is common to gets errors of up to +/- 10% in anemometer surveys of airflow. At other times, equipment moving about, doors opening or shutting, and other temporary mine disturbances can effect airflow measurements.
If the model still gives very different results from actual measurements, don’t be tempted to blame the software or force the software to give correct answers by artificially changing or fixing settings. Provided the simulation is performing without errors or warnings, it is almost certainly simply returning the results of the input data placed within the model. Work backwards from the main airflows to the surface and try to discover the discrepancy between the data put into the model (such as airway sizes, frictions factors, regulators or doors, fans etc), and the actual data of the mine. Some parts of the mine may require a physical inspection – is the fan running? Has the drive or airway been filled with backfill or other blockages? Is the door or regulator open or close?
More often than not, a major factor or item has been missed in the model simulation, and once resolved the simulation will perform as expected.