Optimal Decisions for Forest Operation Managers

1
Optimal Decisions for Forest Operation Managers
The best way to predict the future is to create
it!
1 of 28
2

• The forest operation planning problem is big,
  messy and expensive!
• Sub-optimal plans cost companies millions in
  higher costs and lower revenues
• Even simple, feasible plans can take lots of time
  and effort
• Changed circumstances require a long process of
  re-planning or crude adaptation
• Plans are not checked for robustness, so more
  risk/or less efficiency

2 of 28
3
OperMAX is for

• Multi-Year Operational Planning
• Optimal Annual Planning
• Annual and Multi-Year Re-planning
• Strategic and Tactical Decision-making
• Sensitivity Analyses

3 of 28
4
OperMAX Planning Applications
MIP Optimized Multi-Year Planning Analysis
for Entire Woodlands
-harvesting -transportation -road construction
silviculture -product bucking supply
chain -wood product sales purchases -millyard
inventory management
Strategic Analysis -land acquisition
blocking -fleet management -product woodflow
analysis -sorting yards -capital budgeting
projects
Shadow
Exports for
Exports for
Prices
Optimized Plan for Years 2
Optimized Annual Plan Budget
by months, seasons or user-defined periods
Tracking
Regular (Weekly or Bi-Weekly) Optimized
Re-Planning
Longer-term
based on monitoring of actual operations
-block product volume deviations -wood product
price changes -fire and/or insect
outbreak -equipment breakdown
e.g.
Actuals
4 of 28
5
Our Company

• Helping managers make better decisions, helping
  forest products companies save millions and gain
  a competitive advantage
• Focus on providing integrated solutions for the
  management of operations and support for
  strategic decision-making
• Expertise in forest operations management,
  optimization and IT
• 20 years of experience in projects on three
  continents

5 of 28
6
OperMAX Case Study

• Scenarios were developed based on operational
  situation for a large forest products company in
  eastern Canada
• A simulation of companys current decisions was
  run, then the same data was used for optimization
  runs
• Used data from their blocking (remaining three
  years of the 2002-2007 forest management plan)
• Mill prices, cost factors modified to maintain
  confidentiality, used FERIC productivity
  equations
• Results should be taken as indicative
• Objective Study value of OperMAX in a realistic
  situation

Forest Engineering Research Institute of Canada
6 of 28
7

• The existing 2002-07 plan spelled out
• Block-system allocation what system in what
  block
• Block-season allocation what season for each
  block
• Products were shipped to closest mills
• Heuristic to reduce transport cost
• These designations were replicated in OperMAX
• Used companys existing decisions to allocate
  system-to-block, block-to-season
• Transport was always to closest mill accepting
  that product
• Block-year decisions were all kept open, thus
  plan was still partly optimized

7 of 28
8
Note that the data supplied to us did not include
market information for several species and
products, therefore some products were harvested
but not trucked
8 of 28
9

• Then a basic optimization run was done
• All operating information (productivities, costs,
  prices, mill consumptions, volumes, etc.) were
  kept the same as in the simulation run
• Site-based operability constraints were kept for
  block-season and block-system decisions
• Otherwise everything else (which systems in which
  blocks to produce which products for which mills
  in what seasons and years) was optimized

9 of 28
10

Why should this happen? It can only be because
the OperMAX system found ways to increase revenue
through careful delivery timing (product prices
are higher in some seasons), while keeping
harvesting and trucking costs low
A huge increase in profit margin occurred in
the optimization despite the negligible
differences in harvest trucking cost!

10 of 28
11

• More optimization runs were done
• Analysis of modeling with mill product groups
• Analysis of harvest system policy
• Supply chain modeling
• Shadow price analysis of
• Harvest system configuration
• Mill inventory levels
• Each resulted in a NPV increase of 0.5 to 3
  million

11 of 28
12

• Different bucking specifications
• E.g. Some timber can make small sawlogs or large
  studlogs, depending upon the location of the
  block and the mill requirements at the time the
  block is harvested it is best not handled as a
  deterministic decision
• Varying Product mixes
• E.g. A pulpmill that can accept several hardwood
  species for its volume consumption

12 of 28
13

Lower trucking costs are part of the reason for
these significant profit margin improvements, but
the rest comes from better timing and getting the
best price possible from external markets
13 of 28
14

• Sometimes operational policies are maintained
  without analyzing the cost of the policy
• E.g. Certain contractors are not used in certain
  districts
• Company policy said that MHS-SW MFT-HW systems
  were only allowed in some districts, but not in
  the largest district
• This was modified to allow all systems in all
  districts (although block-specific constraints
  were maintained)

14 of 28
15

Taking away policy-based block constraints saved
1.5 million, primarily due to significantly
lower harvesting costs so the managers need to
ask themselves if the other benefits of the
policy are worth ½ million per year

15 of 28
16

• Often, in a multi-mill context, mills produce
  by-product that can feed other mills, while
  sort-yards are used to distribute products
• Sawmill chips send to a pulpmill
• Products may be sent to a sort-yard or directly
  to a mill
• Incorporating this reality will make the model
  more accurate as this wood flow will impact
  block-mill deliveries and the scheduling of the
  harvest

16 of 28
17

Improving the supply chain can increase revenues
and may decrease costs but note that, in this
case, all improvements came from better revenue

17 of 28
18

• OperMAX provides managers with shadow prices
  for
• Blocks ? blocking and land acquisition decisions
• Machines ? fleet capacity investment issues
• Mill-yard inventories ? inventory management
• Wood purchase levels ? purchase sales contracts
• Mill production levels ? short, medium and long
  term planning of mill expansion, re-engineering,
  planning shifts per day, shut-downs

18 of 28
19
Using shadow prices is a good way of figuring out
how to best allocate your resources. In this
case, the shadow prices suggest to us that the
feller-buncher capacity should be shifted from
the MFT-HW to the MFT-CHIP system types
19 of 28
20

In this case, the small change of equipment from
one system type to another resulted in
significant improvements to profit margin,
primarily through better allocation of resources
to maximize revenue

20 of 28
21
By again studying shadow prices it was possible
to identify that the maximum mill inventory
constraint at the studmill was costing money

21 of 28
22

A second scenario was run with the max mill
inventory bumped up to produce these results,
primarily due to higher revenues

22 of 28
23


Simple optimization produced increased NPV by 12
million, profit margin by 7 million dollars
More detailed analyses produced another 8 million
of NPV and 10 million dollars of increased profit
margin
23 of 28
24
An annual plan, directly based upon the optimal
MYOP but at a finer time resolution (e.g.
monthly) can be easily created After running the
one-year optimization, OperMAX presents all
results. For example, this report monthly
information about delivered volume, consumed
volume, revenue and so on, for each mill
24 of 28
25
By downloading harvest and transport actuals and
entering any other new information, a new plan
can be created taking into account the updated
information The new reports distinguish between
information from actuals and the revised plan
25 of 28
26

• Will enable optimized integration between mills,
  harvest, roads, transport
• Enterprise-wide wood-flow modeling
• Supports strategic analysis capabilities
• Can be used for annual and routine operational
  re-planning
• Therefore has potential to transform the
  operational planning flow, touching all core
  aspects of woodlands
• And in the process make a difference on the
  bottom line, in the millions of dollars per year

26 of 28
27

• This technology is not a question of if but
  when it will be adopted
• Computer power is sufficient, potential benefits
  are huge
• Being on the front-end allows you to develop
  the in-house expertise to retain competitive
  advantage
• Being on the front-end may allow you to
  leverage the risk with RD tax credits
• Given the huge upside potential, its a low risk

27 of 28
28
We are now scheduling onsite and online
demonstrations of OperMAX. Contact us to arrange
yours
Phone [phone] Fax [phone] E-mail
ewr_at_fra.nb.ca URL www.fra.nb.ca
28 of 28