Taking Fluid elements to the next level

Hi,

At the moment, the Fluid elements are well designed, but aimed mostly at fairly basic scenarios (e.g. combine various fluids to make a product). I have some questions about how I could take them to the next level, by extending the basic functionality.

My scenario is Greenhouse Gas modelling in underground coal mining. In this I have a variety of fluid streams, namely:
- waste methane streams, with components like air, water, CH4 and CO2 at different proportions
- emissions streams, which occur after utilisation (burning) of methane and include air, water, CH4, CO, NO2 and CO2 at different proportions
- energy streams, such as diesel or electricity which are used to power machines, and produced secondary emissions
- coal streams, which are volume flows of broken coal with Rosin Rammler (Weibull) particle size distributions, each size distribution with its own methane content which desorbs proportional to particle size and time

In this scenario then, coal mining is driven by comunition (size-change) events, where one large reservoir of coal is broken into successively smaller particle size distributions, by various operations, until the desired size distribution is achieved. The coal all starts out with the same amount of gas in place (e.g. 10 m3/tonne), and the methane starts to desorb as soon as it is exposed to air pressure, and each lump (e.g. 50 mm) will take approx 20 days to completely de-gas.

Various options can be taken to reduce the amounts of emissions, such as drilling holes before mining, sucking out methane and utilising it to make electricity, thereby making profit and reducing the amount emitted post-production. Since the holes are exepnsive to drill (e.g. $1million each), we want to simulate the process to determine the optimal approach to redcing emissions costs (i.e. carbon tax).

In order to solve this problem I have some questions about how to extend the basic Flexsim fluid metaphor to suite complex model:
1. Is there a limit to the number of sub-components that are loaded into the Ticker? Or do I simply list every conceivable element and then make my fluid products out of groupings of elements.
2. How would I go about converting one fluid type for another where the volumes change? as an example consider the burning of methane to create electricty and emissions, where the volume of CO2 produced is much larger than the volume of CH4 burnt (molar ratios), and a secondary fluid (electricity is also produced).
3. How is it best to consider energy used and/or produced, if this needs to be tracked for every operation? Should energy be a fluid, or a property like cost?
4. How is it best to track the particle size distributions, and assoicated methane contents, of the coal as it flows through the flid network? Should I use bucket approach, where the fluid is made out of 10 buckets, and I use global variables at each operation to track the actual sieve size (smallest particle) of each bucket, and simply change these as I move through the network?

I think your product is excellent and would appreciate help in tailoring it for this more sophisticated scenario. Please advise.

Thanks

Brett Forbes
PhD Student
Universit of Queensland

Hi,

Since no one has had the time to write a useful response to what I think are intriguing questions, I'll up the ante.

Consider the conventional use of discrete and continuous simulators by companies, and Flexsim's place in that market. At present much of the work in the simulation market is done by consultants, and only a few companies purchase simulators for in-house use (i.e. limited license sales). Why?

Answer. Partially because of perceived value gaps, such as skill-set, learning etc, but mostly because of the perceived value return, which is not perceived as critical. So the product is a nice to have in the market not a must have.

What is the best way to change this situation? Answer. Define a market segment, where a hybrid simulator is a must have.

As IT systems have evolved over the last 25 years, a gap has appeared in the knowledge management market, a meso-level (i.e. above technical, and below ERP) decision support need has arisen in companies with complex production processes (e.g. mines, oil & gas, some manufacturing). In these environments, a series of very complex and expensive trade-offs need to be made in a sophisticated multi-objective optimisation problem (i.e. more than one right answer), with a lack of effective decision support.

What is needed is a means to model the circumstances and consequences which are a result of the technical details, so that the exact impact can be determined and optimised.

Thus, discrete and continuous simulators have a unique, high value role in interpreting the results of both other types of simulators (e.g. computational) and directly measured results (e.g. SCADA). If a model is defined which maps the underlying process, and in particular the circumstances and consequences of underlying results, then this becomes incredibly useful as a means of examining complex, high value decisions.

In the example application to this thesis, described above, that of a coal mine generating greenhouse gases, which they are forced to track and report on, then it is clear that Coal mines will be prepared to pay for this type of decision support. In addition, there are extensive SCADA systems and a sophisticated simulation environment, including geo-mechanics, reservoir and CFD modelling. Thereby they are a high-value first market segment with a real, high-value need.

So in summary I note, that my question above pertains not just to a new series of extensions to Flexsim to suit the application, but more importantly creates a new type of decision support tool to suit an entirely new and uncontested market space, as well as pointing the way for other similar extensions.

I look forward to discussing these ideas further.

Regards

Brett