From jpretti@barrow.uwaterloo.ca Fri Feb  6 09:41:32 1998
Date: Wed, 4 Feb 1998 09:31:17 -0500 (EST)
From: John-Paul Pretti <jpretti@barrow.uwaterloo.ca>
To: hwolkowi@orion.math.uwaterloo.ca
Subject: Case Study 1


Here are my comments to the students.

The case studies were done fairly well. I used the given case
study/project evaluation scheme to as a guide in marking. In very general
terms this is how marks were allocated to this particular study with areas
resulting in lost marks explicity mentionned :

Communication (20)
 - summary
 - conclusions
 - appendices i.e. GAMS program and output included
 - GAMS documentation
 - organization, clarity, self-containment

GAMS (15)
 - used
 - highlighted or annotated
 - use of scalars
 - listing of units for data

Extensions (5)
 - at the very least, some mention of how problem can be extended
 - at the very best, attempt an extension of the problem
 
Sensitivity Analysis (20)
 - explanation of how results achieved
 - discuss an increase in CN (explicity mentionned in problem)

Discussion of Model (20)
 - list a number of assumptions (e.g. rounding, estimates, waste,...)
 - judge relevance and accuracy of results

Solution of Problem (20)
 - sets
 - data
 - eqns (demand, fractionater and cracker capacity,...)
 - give and explain objective function

Some further notes:

1. CHECK THAT YOUR SOLUTION IS REASONABLE !!! You should not get a
solution yielding > 100,000 barrels

2. Many ignored or modified the cracker capacity constraint
 
3. If you included variables for both input and output for the
fractionater and cracker than you shouldn't need to do any calculations by
hand.

4. I wrote on several studies that it helps me to see the entire problem
statement. It is okay if it is scattered throughout the DATA and other
sections as long as it is easy to follow and find.

5. If you have major technical difficulties and can't get a reasonable
solution, demonstrate that you have a full understanding of the model to
try and recoup some marks.

The study will be discussed in class. Please see me if you still have any
questions at all about the marking or problem. The more feedback I get,
the better job I can do at helping you prepare for the project.


John-Paul Pretti (J.P.) 
University of Waterloo
Department of Combinatorics & Optimization
MC 5136A (519)888-4567 x6895