Pipe networks can be simple or complex.  That seems like an obvious statement, but I probably don’t mean what you initially thought I meant.  At one time, I was working with software which was basically a modelling toolkit.  This toolkit could be used to provide customised software which would model specific types of networks such as telecommunications networks or electricity supply networks or any other network for which a data model had been specified.

At the time, I was trying to configure the software to model natural gas networks.

I went through the basic steps of producing a data model which could represent a network.  That much was easy.  But then I had to decide what characteristics of pipes and their arrangement were important enough to include in the model.  Pipe diameter and length seemed important, and obviously the location was vital, but what about the pipe material?  Was the wall thickness important?  The slope of the pipe?  Depth of cover?  Age?  Corrosion protection?  Price?  National or international standards each pipe might satisfy?  The list of questions can be endless and for each, some modelling decisions must be made.  I could make a simple model with limited detail or a complex model which stored detailed information about every possible item.

The problem was that I knew very little about natural gas networks.  I knew that gas pipes could have very low pressures and that traps were needed to collect condensation from the gas, but when it came to deciding what information was essential to model a gas network productively, I was lost.

In the end, I had to give up: I couldn’t tell what to include and what to omit.

The lesson?  To be successful in data modelling, you need to understand the things you are modelling.  You also need to know exactly what needs you are trying to satisfy.

 

Data Modelling that Works

Our new course Data Modelling that Works examines practical ways of balancing completeness and “data overwhelm” as well as many other aspects of data modelling.  Enrolments are now open. For more information on the curriculum or to sign up, please visit:
http://online-courses-with-mark-t-morgan.thinkific.com/courses/data-modelling-that-works