Overall Project Review - Paper Furnish Optimization3
1. Lean Six Sigma
Paper Furnish Optimization
Black Belt Project
United States Gypsum Co.
Project Launched: 1/27/10
2. Define
2
• Value Stream Map for Deeper
Understanding and Focus
• Identify Key Input, Process and
Output Metrics
• Develop Operational Definitions
• Develop Data Collection Plan
• Validate Measurement System
• Collect Baseline Data
• Determine Process Capability
• Complete Measure Gate
Improve
Lean Six Sigma
DMAIC Improvement Process Road Map
• Project Charter
• Voice of the Customer
• SIPOC Map
• RACI Chart
• Communication Plan
• Effective Meeting Tools
• Inquiry and Advocacy Skills
• Time Lines, Milestones,
and Gantt Charting
• Review Project Charter
• Validate High-Level Value Stream Map and Scope
• Validate Voice of the Customer
and Voice of the Business
• Validate Problem Statement
and Goals
• Validate Financial Benefits
• Create Communication Plan
• Select and Launch Team
• Develop Project Schedule
• Complete Define Gate
ToolsActivities
ControlMeasureDefine Analyze
Kaizen, 5S, NVA Analysis,
Generic Pull Systems,
Four Step Rapid Setup Method
Identify and Implement Quick Improvements with Kaizen
3. Define
3
Project Charter
Problem: Optimization of furnish storage, blending, and
processing has not been realized.
Uncontrolled furnish blends lead to added costs for:
using more expensive furnish when not necessary
Goal: Optimize furnish handling and processing to achieve a
paper cost reduction of $xx,xxx (using expected furnish
pries from Jul10-Dec10)
Furnish substitution in filler system
Problem/Goal Statement
Gate Review Schedule
Financial Impact
Team
Reduce paper cost $xx,xxxc versus 2010
Furnish pricing using Jul10-Dec10
Total potential savings of $xx,xxx
All $ at this point are Type I hard savings
Will address weight reduction opportunities in Project 2.
Core Team members (% of contribution) – 95%
Supporting members – 5%
Project sponsor – Plant manager
LSS Deployment Champion – Corporate
Black Belt
Master Black Belt coach
Tollgate Scheduled Revised Complete
Define: 02/26/10 - 02/26/10
Measure: 03/26/10 04/01/10 04/01/10
Analyze: 04/23/10 06/18/10 06/18/10
Improve: 05/14/10 03/24/11 03/24/11
Control: 06/04/10 04/25/11 Scheduled
4. Define
Project Operational Definitions
Furnish/wastepaper
Bales of recycled paper “scraps” that are used to make gypsum wallboard
paper, which consist of layers, or “plies”.
Gypsum wallboard
Generic name for Sheetrock (USG Trademark product).
Liner/Filler
Liner refers to the top several plies of paper, i.e. the “face” of the sheet.
Filler refers to plies of the paper that lie underneath the liner plies.
Pulper
Equipment that mixes the wastepaper bales with water to form a stock or
slurry of paper.
Felt
Continuous pad that plies of paper are sprayed onto, provides support for
sheet of paper until it is thick enough to travel through equipment.
4
5. Define
Project Operational Definitions (cont.)
Filler Cobb/Liner Cobb
Quality test that determines how many grams of water the finished sheet
of paper (filler side or liner side) will absorb in a given amount of time.
Tensile
The strength of the paper when measured by how much force it takes to
pull apart a strip of paper (in the MD or CD direction), another key quality
test.
MD/CD
Machine direction (parallel) or Cross direction (perpendicular), refers to the
orientation of the paper fibers in the sheet.
Basis weight
The weight of the paper in a given amount produced, measured in lbs per
thousand square feet (#/MSF).
5
6. Define
Project Operational Definitions (cont.)
Furnish blend
Refers to the recipe used in the filler pulper. How many bales of OCC to
how many bales of DLK.
OCC
Old corrugated cartons, wastepaper typically from recycled cardboard
boxes, much of this material comes from landfill sorting facilities.
DLK
Double-lined kraft, wastepaper typically from clippings on new cardboard
boxes, shipped directly from box manufacturers to the plant.
Size
Chemical added to help the paper (when dry) resist the tendency to
absorb liquid.
Silo stacking
Process of stacking the paper rolls in an upright column or “silo”.
6
7. Define
Project Operational Definitions (cont.)
Bale clamp truck
A forklift with a specially designed attachment to grab and lift rectangular
bales of wastepaper.
Furnish bale tags
Informational labels printed out and applied to each wastepaper bale as it
enters the plant.
WWP
Wastepaper Warehouse Plan, an internal document developed to contain
all major procedures involved in the receipt, unloading, storing, and
retrieval of furnish bales at the plant.
5s
Lean Six Sigma process/tool designed to help properly clean-up and
organize any area to ensure efficient use of the space.
7
8. Define
Project Operational Definitions (cont.)
Newsline
Grade of gypsum wallboard paper. Is the brown-colored sheet on the
back of the gypsum wallboard, the side that is mounted to the wood/steel
studs in building construction.
C/O
Changeover. Refers to the process of changing from one product to
another on the paper mill production line.
8
15. Define
15
Team Effectiveness
Formal goals, agenda, meeting minutes, action item lists, and time
schedule used for all meetings
Ground Rules set at Team Launch
Team Roles – assign timekeeper, note taker, and “engager” at each
meeting
Team Meeting Frequency – Each Wednesday, 1-3pm
Tools used by Team – SIPOC, Voice of Customer analysis, Risk
analysis, Communications plan, Project plan, RACI chart
Team has agreed to have an open mind and communicate thoroughly
in order to avoid scheduling conflicts
16. Define
Voice of the Customer (VOC) – Paper Furnish
Optimization
Customer
Voice of the
Customer
Key Customer
Issue(s)
Critical
Customer
Requirement
Who is the Customer? What does the
customer want from us?
What does the customer
want from us? We need to
identify the issue(s) that
prevent us from satisfying our
customers.
We should summarize key issues
and translate them into specific and
measurable requirements
Machine Tender
(Paper Mill Wet End
Operator)
Stock that meets
all applicable
specs
Freeness, consistency,
tensile, stock flow,
good recycled water
quality, minimal
wax/glue/trash
1. Freeness & consistency
within established
operating ranges
2. Tensile strength within
bulletin range for paper
grade
3. Continuous flow of stock
4. Recycled water will pass
visual inspection of
save-all pads
5. Minimize wax/glue/trash
contamination
Wet End (Paper
Machine)
Same as above Same as above Same as above
18. Risk Analysis – Paper Furnish Optimization
• Cover most important risks (80/20 rule) – Top 3
Risk# Risk Description
Impact
Type
Probability
of
Occurrence
Rating
Impact
Rating
Priority
Rating
Risk Response
Owner1 - None 1 - None
(Probability
x Impact)
Accept
Identify and Describe
Risk Response
Scope 2 - Low 2 - Little Avoid
Cost 3 - Medium 3 - Moderate
Mitigate - Minimize
Probability
Time 4 - High 4 - Heavy Mitigate - Minimize Rating
1*
Increased pulping time (pulper
operators weigh every
bale) C 4 4 16 Avoid
Be wary of this.
Perform careful
trials. Foreman
2*
Bad quality finished paper (using
wet/degraded bales) C 3 4 12
Mitigate – Minimize
Probability
Don’t rush to use all
bad furnish at
once
PM
Foremen
3*
Bad quality finished paper
(changing bale blend % i.e.
increasing OCC, etc.) C 3 4 12
Mitigate – Minimize
Probability
Ensure scavenger &
screening
systems are
functioning
optimally
PM
Foremen
4
Increased furnish price ($/ton) by
using local vendors to lower
stock level C 3 3 9 Accept
Team must prove this
cost would be
offset by a
quantifiable
issue decrease Buyer
5
Screening capacity to handle
“dirtier” furnish C 3 3 9 Mitigate – Minimize Probability
Ensure screening system
are functioning
optimally
PM
Foremen
6 Run out of waste paper furnish C 2 4 8 Avoid
Don’t let stock levels fall
too low Buyer
19. 19
Communication Plan – Paper Furnish
Optimization Project
Project
Sponsor
•Face to face
meetings
•Gate review
presentations
•Project
progress
updates
•Go/No Go for
project phases
•Updates and
potential
roadblocks
•Phase review
and validation
•Black Belt
•Project Team
•Weekly
•End of each
phase per
project plan
•Scheduled
•Define gate
review
scheduled
Process
Owner
•Copy of team
meeting
minutes
•Gate review
presentations
•Monitor
project
progress
•Major project
updates
•Overall team
progress
•Phase review
and validation
•Black Belt
•Project Team
•Weekly
•End of each
phase per
project plan
•N/A
•Define gate
review
scheduled
Plant
Employees
•Newsletter
articles
•High level
summary to
keep all
employees
informed
•General “high-
level” items to
communicate
project status
•Project Team
(will rotate)
•One article
generated per
project Phase
Paper Mill
production
and supply
chain
employees
•Shift safety
huddles
•PM
lunchroom
monitor
•Stimulate
employee buy-
in and interest
•Project goals,
benefits
•Effects of
furnish on
paper machine
•Project Team
(will rotate)
•PM General
Foreman
•Twice/monthly
•Twice/monthly
Audience Media Purpose
Topics of
Discussion/
Key Messages
Owner Frequency Notes/Status
20. Measure
20
Measure Improve
Lean Six Sigma
DMAIC Improvement Process Road Map
• Value Stream Map
• Operational Definitions
• Data Collection Plan
• Statistical Sampling
• Measurement System Analysis (MSA), Gage R&R
• Control Charts
• Process Capability, Cp & Cpk
• Identify Key Input, Process and Output Metrics
• Develop Operational Definitions
• Develop Data Collection Plan
• Validate Measurement System
• Collect Baseline Data
• Determine Process Performance/Capability
• Validate Business Opportunity
• Value Stream Map for deeper understanding & focus
• Quick Wins (Control Plans)
• Measure Gate Review
ToolsActivities
ControlDefine Analyze
Kaizen, 5S, NVA Analysis,
Generic Pull Systems,
Four Step Rapid Setup Method
Identify and Implement Quick Improvements with Kaizen
Measure
22. Measure
22
Data Collection Plan – Page 1 of 3
Performance
Measure
Operational
Definition
Data Source and
Location
How Will Data Be
Collected
Who Will Collect
Data
When Will Data Be
Collected
Sample Size Stratification
Factors
How will data be
used?
Stock supply
operator (bales
being used)
Stock supply –
as bales are fed
PM S/C
unloading
operator (bale
age)
PM S/C – daily,
next morning after
(3) shifts worth of
production
Stock supply
operator (bales
being used)
Stock supply –
as bales are fed
PM S/C
unloading
operator (bale
wgt)
PM S/C – daily,
next morning after
(3) shifts worth of
production
Freeness
Units unknown
(plant personnel
do not use units)
Flowmeter
reading logged
into local PM
InSql Server
Logged when
running by PLC
PLC
PLC logs every
10 sec, will look
at every min for
Minitab evaluation
1/1/10 - 3/22/10 Grade of paper
Used as a
proactive
measure of sheet
strength
Consistency
% of stock to
water (3% = 3%
stock, 97%
water)
Flowmeter
reading logged
into local PM
InSql Server
Logged when
running by PLC
PLC
PLC logs every
10 sec, will look
at every min for
Minitab evaluation
1/1/10 - 3/22/10 Grade of paper
Used to
proactively
measure density
of sheet
Tensile (Parallel)
lbs of force (to
tear sample)
Thwing Albert
machine feeds
into MDIS
system, in PM
Quality Lab
Samples of each
set are prepped
and fed into
Thwing Albert
PM Quality
testers
As each set of
paper is produced
1/1/10 - 3/22/10
Grade of paper,
production shift
Used to quantify
strength of sheet,
minimum reject
limits
Data Collection Plan - Page 1
Grade of paper,
Look at filler
system only
Give idea on age
of bales versus
paper grade and
variability
Weight of bales
being fed into
pulper
Weight (lbs)
Unloading log,
PM S/C
unloading office
Record number
of bales being
used (whse
section & row),
then research
unloading log.
Entire run (~ 8 to
9 days)
Grade of paper,
Look at filler
system only
Give idea on
weight of bales
versus paper
grade and
variability
Age of bales
being fed to
pulper
Time (days)
Unloading log,
PM S/C
unloading office
Record number
of bales being
used (whse
section & row),
then research
unloading log.
Entire run (~ 8 to
9 days)
26. Analyze
26
Improve
Lean Six Sigma: Analyze
Improvement Process Road Map
• Cause & Effect Analysis
• Hypothesis Tests/Conf. Intervals
• Simple & Multiple Regression
• ANOVA
• Identify Potential Root Causes
• Reduce List of Potential Root Causes
• Confirm Root Cause to Output Relationship
• Estimate Impact of Root Causes on Key Outputs
• Prioritize Root Causes
• Complete Analyze Gate
ToolsActivities
ControlMeasureDefine Analyze
Kaizen, 5S, NVA Analysis,
Generic Pull Systems,
Four Step Rapid Setup Method
Identify and Implement Quick Improvements with Kaizen
27. Analyze
27
Current Process Performance –
Newsline MD (Parallel) Tensile
Observed Special Causes:
Individual values look good, one special cause in the range
chart that could not be tied back to an observable event
Five products made over observation window (~ 2 days)
28. Analyze
28
Current Process Performance –
Newsline CD (Cross) Tensile
Observed Special Causes:
A minor issue on the individual values, but range chart ok
Five products made over observation window (~ 2 days)
29. Analyze
29
People
Potential Root-Cause Identification (sample)
Measurement Parts
Machines
Why Are Cycle
Times >18 hours
Language barriers
Software Glitches
Unfamiliar Processes
Customer Delays
Lab Personnel
Lack of Experience
Approval required in some situations
Customer responsible for consumable ordering
Methodology
Operational Procedures
SAP System
Call not closed at time of Repair
Unavailable, system down
Parts Unavailability
Operations
Left open for testing
No access to system
Slow, Times Out
Deployment delay
Vendor
Database
Missing Information
Erroneous Information
Parts Order Processing
Part procurement process is substandard
Centralized Call Process not being followed
Constant Process
Changes
Lilly Staff Level
Customer unwilling to add another Tech to Contract
No Backfill for PTO
Order fulfillment Delays by the manufacturer
Corporate decision not to stock parts locally
Excessive Part ETA times
Part Numbers not in
Database
Quality of Parts
Multiple DOA Parts
Recycled parts
Wrong part sent / Published part # incorrect
Lab Technician Off-site
Remote area of Plant
Tech too busy with callsLilly
Unknown escalation path
Inconsistent process
application
Training issues
Manufacturers
Long hold times
Inept technical assistance
34. Analyze
34
Summary of Critical x’s
Critical X1 = Basis weight (to tensile)
Tool used = single regression
P-Value = 0.000
Critical X2 = furnish blend (to Liner Cobb)
Tool used = single regression
P-Value = 0.000
Critical X3 = bale storage rotation plan
Tool used = fishbone diagram, C&E matrix
Will pilot changes while watching BW & tensile variation
35. Analyze
35
Other LSS Tools Used
Normality testing
Distribution identification of data set (more accurately
depict process capability of some key outputs)
Sample Size Calculation (Using sample std dev. and power)
Correlation tools (single & multiple regression models)
Team Effectiveness Tools (Brainstorming)
36. Improve
36
Lean Six Sigma: Improve
Improvement Process Road Map
• Design of Experiments (DOE)
• Solution Selection Matrix
• Piloting and Simulation
• Develop Potential Solutions
• Evaluate, Select, and Optimize Best Solutions
• Develop and Implement Pilot Solution
• Confirm Attainment of Project Goals
• Complete Improve Gate
ToolsActivities
ControlMeasureDefine Analyze
Kaizen, 5S, NVA Analysis,
Generic Pull Systems,
Four Step Rapid Setup Method
Identify and Implement Quick Improvements with Kaizen
Improve
37. Improve
37
Brainstorming, Solution Selection Matrix (Tribal Knowledge)
Used fishbone & CE matrix to develop list of significant inputs
Warehouse organization & bale feed procedure consistency
Critical “foundation” organizational work done. Set a solid
foundation for the team to build any repeatable and consistent
practices on top of. Many benefits realized in the form of job
efficiency, employee morale, consistency in procedures from
shift-to-shift, etc.
Team felt getting organized (alone) would help improve MD
tensile variation…….. it did NOT (see next slides)
Most other inputs did not show significant correlation to MD or
CD tensile results
Potential Solution Identification
40. Improve
40
Pilot Plan
DOE
Part I
• First part of a two-part DOE newsline trial.
• Part I used “clean” OCC furnish from Interplast.
• Hold critical non-DOE
factors stable
• Complete required set
point combinations
• All members • Completed
1/26/11
DOE
Part II
• Second part of a two-part DOE newsline trial.
• Part II used “dirty” OCC furnish from Vista Fiber.
• Hold critical non-DOE
factors stable
• Complete required set
point combinations
• All members • Completed
2/24/11
DOE
Combining
• Had to first use statistical (mean/median) testing to
ensure the (21) critical variables remained “the same”
during both trials so results could be combined together
into one DOE.
• All critical variables were
the “same”, considering
the practical aspect of the
operation
• Black Belt • Completed
3/4/11
DOE
Analysis
• Complete DOE Analysis for MD and CD Tensile. • Statistically solid results
after analysis is complete
• Black Belt • Completed
3/8/11
ScheduleTeamSuccess CriteriaDescription
Pilot
Test
44. Improve
44
Quick Wins (6)
Forklift backhaul procedure
Lower occurrence of empty backhauls
Paper bale use from warehouse
Ensure consistent shift-to-shift use strategy
Section A “silo” stacking
Received approval from FM Global, 48 roll capacity vs. 36
Paint striper
Good quality tool that allowed us to efficiently paint indoor/outdoor
storage lines for wastepaper bales
Bale clamp trucks overheating
Developed OBC (operator basic care routine)
Furnish bale tags
Still in process but will yield powerful information
45. Improve
45
Current Status Summary
Key actions completed
5 of 6 identified Quick Wins completed
Implemented key sections of WWP
Began tracking filler pulper furnish blend
Installed furnish blend board at pulper
Designated indoor storage for consistent use of outside furnish
Built two “significant inputs” matrices after thorough analyzing of
multiple production runs
Began labeling all incoming DLK & OCC bales
Completed DOE
46. Improve
46
Current Status Summary - Continued
Lean Six Sigma Tools used
5s in raw paper warehouse (storage for furnish bales)
Solution selection matrix
DOE (Design of Experiments)
Lessons learned
Some “Quick Wins” may not be so quick, but still important!
Scope on this project is very large
Some results are not crystal clear, must sift through process
“noise”
Trials are difficult in low business environments (delayed project ~
three months)
47. Control
47
Lean Six Sigma: Improve
Improvement Process Road Map
• Mistake-Proofing/Zero Defects
• Standard Operating Procedures
(SOP’s)
• Process Control Plans
• Visual Process Control Tools
• Solution Replication
• Project Transition
• Implement Mistake Proofing
• Develop SOP’s, Training Plan and Process
Controls
• Implement Solution and Ongoing Process
Measurements
• Identify Project Replication Opportunities
• Complete Control Gate
• Transition Project to Process Owner
ToolsActivities
MeasureDefine Analyze
Kaizen, 5S, NVA Analysis,
Generic Pull Systems,
Four Step Rapid Setup Method
Identify and Implement Quick Improvements with Kaizen
Improve Control
50. Control
50
The following SOP’s and/or Documents have been updated,
tested, and implemented:
Throughout the project, (15) documents were
created/revised. These were various standard operating
procedures, visual work instructions, equipment operating
manuals, etc.
Standard Operating Procedures (SOP’s)
and Documentation
51. Control
51
Training Plan
A training plan was created for each item covered in the SOP and
documentation section slides shown previously.
A list of training observations and “training lessons learned” was
also created.
53. Control
53
Replication Opportunities
A list of replication opportunities was developed. This mainly focused
around other USG facilities that could benefit from the findings in this
project.
55. Control
55
Project Risk Management
Three main risk areas were brainstormed and risks identified for each
major category listed below.
Business Risk Management: Potential risks resulting from the
planned solution and implementation plan on the business and its
customers
Organizational Acceptance Risk Management: Risks related to
lack of cultural buy-in for the proposed solution and implementation
Sustainment Risk Management: Risks related to sustaining the
project gains over the long-term
57. Control
57
Lessons Learned
Based on learnings from this project:
What would you have done differently?
Better use of technical resources (bale labeling/scanning)
Possibly divided this project into pieces (very large scope)
What worked really well?
Team selection/balance worked very well
Engaged employees through early buy-in of objectives
Use of visual controls (furnish board, signposts, c/o time board)
What will you do differently in future projects?
If data results are variable, look to use weighted selection matrix
sooner in project
58. Control
58
Lessons Learned (cont.)
Based on learnings from this project:
What are some additional recommendations related to
lessons learned?
Schedule
Tough to meet initial schedule with large project scope
Trial scheduling was difficult (low news volume, production
schedules)
Business Impact
With decent pricing spreads between the various furnish grades
there is money to be saved!
Teamwork
Took the talents of the entire team to keep project moving
Discovered talents of other personnel in department as well
59. Control
59
Financial Benefits
Considering furnish substitution
Charter proposed savings of $xx,xxx/yr
Actual 8 month savings = $xx,xxx ($x,xxx/mo, $xx,xxx annualized)
Introducing new procedures (not yet optimized) earlier in project most
likely led to reduced attainment of potential monthly savings
Remaining difference tied up in real blend increases seen
Charter goal proposed increasing type x furnish blend% by 25%
Actual “end of month” increase achieved has only been 5%
Lower newsline volume, short run cycles, and changeover ramp time has made
it difficult to achieve the END OF MONTH blend goal of a 25% increase
Actual blend % when running newsline products is a 50% increase!!
Mar11 savings = ~$x,xxx (17% type x furnish increase), C/O times
being reduced
60. Control
60
MGPP
Multi-Generation Project Plan (MGPP)
Generation 1
2010/2011
Generation 2
2011
Generation 3
2012/2013
Vision Use current LSS team to
improve furnish handling
process.
Optimize wet end chemistry in
order to reduce shrinkage,
improve drainage, and lower
size usage.
Install (2) xxx formers to
better control basis weight
variation.
Process
Generation
First major step to be
completed. Will focus on
operating procedures with
minimal capital investments.
Second step after successful
completion of furnish
optimization project. Will
include vendor on team and
require minimal capital
investments.
Third major step, will require
approximately $1,000,000 in
capital investments.
Platforms /
Technology
Use current MDIS system and
PLC data. Also uses manual
documentation sheets.
Use current MDIS system and
PLC data. Also uses manual
documentation sheets. Vendor
will likely need to supply data or
additional instrumentation for
critical variable monitoring.
Will use PLC and other
similar technology. This
forming technology is much
better at producing a
consistent and uniform
sheet.