Question on nonconserved flows

["Prof. Dr. Niall Palfreyman"]

Hello,

part of my work involves the mapping of chemical reactions, and there a
phenomenon often crops up which is only awkwardly representable in Stella.
To take an example: let us suppose that 100 atoms of oxygen interact with
200 atoms of hydrogen to produce 100 atoms of water. On the face of it
this means that 300 atoms of hydrogen and oxygen disappear, and only 100
atoms of water appear - the reaction is a nonconservative flow.

Now before anyone sets me straight on this, I am of course aware that
there are a number of ways of modelling this in system dynamics. I can
model atomic weights instead of numbers of atoms, in which case the amount
of matter is conserved in the reaction, and so can easily be modelled
system dynamically. Or I can use coflows to model the reaction.

There is however a definite problem with these modelling methods in that
none of them really represent the reaction the way a chemist sees it, ie,
in terms of numbers of atomic configurations which dont just appear out
of clouds, but are really juggled around structurally. For the chemist a
reaction really is a nonconserved flow which cannot be convincingly be
represented as conserved flows.

Now my question is this: Is this a phenomenon which only arises in
chemistry, or are there other areas in which a similar situation arises? I
would be very interested in hearing about _any_ other concrete and
practical situations in which flows are (a) nonconservative and (b) not
convincingly represented by loads of mythical clouds.

Best wishes,
Niall Palfreyman.
From: "Prof. Dr. Niall Palfreyman" <niall.palfreyman@fh-weihenstephan.de>

["Mark B. Wallace"]

If you are going to create models of chemical reactions, it might be useful to learn a
little bit about chemistry first. There is no such thing as an "atom of water." Water
(H2O) is a molecule, not an atom. In your reaction, 300 atoms (of hydrogen and
oxygen) combine into 100 molecules (of water). There is no conservation problem.

The more general lesson from your post is that SD skills, alone, are not sufficient to
create a useful model in a subject area about which the SD practitioner knows
nothing. In such cases, the SD expert should either (1) learn the subject matter, or
(2) teach SD to someone who is a subject matter expert, and let them create the
model, or (3) work collaboratively.

From: "Mark B. Wallace" <mark.wallace@verizon.net>

[James Winebrake]

Not sure if this helps, but...

Ive used STELLA for modeling chemical reactions. I have looked at the
process as one of unit conversion, as opposed to non-conserved flows. In
effect, the units of the system (i.e., atoms) are converted to other units
(i.e., molecules). We had a model we designed for a chemistry course which
illustrated limiting factors laws. We used moles instead of atoms, but the
idea is the same...we applied a unit conversion system to model this.

Similar examples abound in other fields. In particular, manufacturers have
to deal with this. Suppose I build chairs. I use planks of wood and nails. I
have a stock of 1000 nails (e.g., hydrogen) and I have 250 planks of wood
(e.g., oxygen). How many chairs (water) would I be able to produce?

In the chemical system we modeled, we allowed the user to change the
chemical equation coefficients (i.e., the molar ratios). Then we used these
numbers to dynamically adjust the unit conversion factors.

Let me know if this model would be of use and I will try to dig it up.

James Winebrake

____________________________________________
James J. Winebrake, Ph.D.
Associate Professor and Chair, Public Policy Department
Director, University-National Park Energy Partnership Program
Rochester Institute of Technology
Rochester, NY 14623
(p) 585-475-4648/ (f) 585-475-2510/ (e) jjwgpt@rit.edu