Introduction
The company
has a conveyor track powder paint line system that allows parts to be painted
to be hung on a conveyor track and these then pass through a 5 stage
pre-treatment wash, then a drying oven before being powder coated and cured in
a curing oven. Approximately 3 years we took on a new customer that has doubled
our paint line production but has almost tripled the amount of pre-treatment
chemicals used. This is generating a large amount of phosphate pollution that
our current installed waste water treatment facility is struggling to cope
with. It is proposed that changing to greener chemical solutions will reduce
the financial cost and environmental impact, while still maintaining all
customer requirements.
Define
The first
step of any Six Sigma project is the Define stage. This is carried out to
determine what the issues related to the project are, the scope of the
investigation and what goals are to be achieved.
Project
Charter: The team developed the Project Charter through discussion to give the
project a clear direction through the Mission Statement. This was then expanded
in the Problem Statement that focused the team to explore why this project
needed to be tackled and what the base line metrics were. Following this
natural progression the next stage was to set the goals that were hoped to be
achieved. This required the collection and analysis of historical accounts to breakdown
the costs into the separate areas in question.
Project
Schedule: A Gantt chart was developed to give deadlines for the various stages
of the project. The time frame for each phase is relatively short so this chart
will ensure that all team members are aware of when individual sections are to
be completed.
SIPOC Chart:
The SIPOC chart was developed as a means to better understand the High Level flow
of parts through the entire manufacturing process. Special attention was paid
to the pre-treatment stage to ensure that all the team members understood what
was happening to the parts during this process, which is the change in the
surface structure of the material that ensures better adhesion of the paint to
be applied. This has to achieve two Critical to Quality features that our
customers demand. The first is the quality of finish on the final painted parts
that must be free from defects or discolouration. The second is the measure of
the performance of the paint through a number of destructive tests and the
completion of 750 hours of Salt Spray testing as carried out in accordance with
ISO 9227.
Fig 1: SIPOC Chart –
High Level Process Flow
|
Measure
The purpose
of the Measure phase is to establish base line data for the process as it is in
its current state. With this in mind the project team set about collecting the
relevant metrics that are associated with the pre-treatment process. Using the
Project Schedule developed on a Gantt chart the first step was to establish how
much was spent on pre-treatment chemicals and how much was spent on the
maintenance, removal and treatment of used chemicals.
This was
broken down into the following areas.
Total spend on the three different
chemicals used in the pre-treatment process – This includes the regular
chemical adjustment to ensure correct active levels in the tanks and the annual
complete replacement.
In House costs – Including time spent
ensuring correct chemical concentrations, time spent cleaning the spray nozzles
of sludge build up, time spent draining and treating the wash tanks, gas costs
for running burners for heating chemical tanks and annual shutdown cleaning.
Treatment costs – Including the
annual cost of IBC’s bought to empty the chemical tanks and to have these used
chemicals taken offsite and treated by a third party.
Collecting
all this data was time consuming as, although most of this was being recorded,
it had never been collated and analysed as an entire process.
The process of examining the various
inputs and outputs of the pre-treatment line provided a good basis to develop
the Process Flow diagram. This is a higher level version of the SIPOC chart
developed in the Define phase and captures the Pre-treatment process with all
its inputs and outputs. Through the development of this diagram the team
developed a better understanding of where the costs attributed to the process
were being consumed and gave better focus to how the improvements being
considered would impact these costs.
As the purpose of the paint line as a
whole is to provide our customers with a finished painted product that is to
the requirements they have stipulated it was decided that a VOC/CTQ chart be developed
for the Measure phase. This is shown below and highlights all the CTQ
requirements that must be achieved for all painted finished goods leaving the
factory. All of the tests are carried out in house with the exception of the
Salt Spray Resistance which is outsourced to several different specialists so that
results can be cross referenced.
Fig 2: VOC/CTQ Chart
Analyse
With the Measure phase completed and
all the base line metrics gathered we could move on to the Analyse phase of the
project.
A meeting was scheduled for all the
team members to come together to brainstorm the causes and effects assigned to
the problem being investigated. This was developed through an Ishikawa diagram.
The framework of the diagram was drawn on a white board and the problem
statement written at the head. This ensured that all team members were focused
on the problem at hand and did not drift off track too much. It was decided
that we should use the 6M categories and the causes of the problem explored for
each of these categories (see Fig. 3 below).
Once the Ishikawa diagram was
complete it looked as though we had the main causes of the problems identified.
To follow the natural progression of the line of thinking the Ishikawa
brainstorming session developed we decided to highlight the main causes and
investigate why these were happening. The tool that lent itself perfectly to
this process was the 5 Why analysis.
Due to the fact that the project is
not investigating product failures but rather increasing costs, both financial
and environmental, it was decided during a meeting to break the 5 why analysis
into 4 sections. These were 2 financial costs, maintenance costs and heating
costs, and 2 environmental costs, waste chemical treatment and waste water
treatment. Each of these was explored and gave the team more focus on areas
that should be able to be explored in the Improvement phase.
The team also wanted to analyse the
financial data that had been collected in the measure phase. It was decided
that the best way to understand the breakdown of cost was to present the data
in a Pareto chart. This was developed and it clearly showed the gas cost of
running the pre-treatment tank heaters and the cost of purchasing the chemicals
that are used in the tanks accounted for over 80% of the annual spend. It is
understood that it will not be possible to eliminate all of these costs but it
is hoped that significant reductions can be made in one or both of these areas.
The requirement to send waste
chemicals off site and the IBC’s needed to transport the chemicals make up the
next 12% of the Pareto chart and is an area that we will be targeting with the
aim of significantly reducing this cost.
A meeting has been scheduled with the
chemical suppliers to discuss the findings made by the team. Some good initial feedback
has been received with several potential alternatives to current chemicals used
and line set up that will be reviewed in more detail at the meeting. A site
visit has also been scheduled to another firm that has been using alternative
chemicals to produce the pre-treatment required. It is intended to have some sample
panels prepared on their line that we can paint and then subject to our test
procedures to determine the effectiveness.
Fig 3: Ishikawa Diagram
Improve
Having Measured and Analysed all the
data we had collected the next phase of the project was to Improve the process
to reduce the costs and phosphate production. Our chemical suppliers arranged a
site visit to a firm that has been using the proposed pre-treatment system for
some time. We were very well received and allowed to ask questions relating to
phosphate and sludge generation and cleaning schedules that backed up the
chemical suppliers claims. Unfortunately this firm did not have the same level
of CTQ requirements for paint finish performance so several aluminium and steel
chips were brought along and allowed to be run through there pre-treatment
process. These were then painted on our system with a selection put aside for
destructive testing and a selection was sent away for salt spray testing.
The teams preferred system of
exchanging the phosphate conversion coating chemical to a zirconium based
equivalent, would require the entire pre-treatment line to be shut down and
completely cleaned through before the setting up of the new configuration could
be carried out. This will take several days to complete and once the new system
is in place it will need several more days to balance the chemicals and ensure
the process is working correctly. For this reason it was decided to carry out
the changeover during the annual factory shutdown at the end of July and early
August.
The team had data from the chemical
suppliers that showed that we can expect due to the very low levels of
phosphate in the new chemical the time spent treating the waste water could be
reduced by up to 50%. Also the sludge generated by the new chemical would be
60-70% less than the current chemical which would equate to 75% reduction in
maintenance time removing and cleaning the spray nozzles.
It has also been shown that the
projected annual usage of the new chemical should be 50-70% that of the current
phosphate based chemical. When taking into account the change in conversion
chemical and the complete removal of the passivating chemical from the last
stage we can expect a reduction in the chemical inventory of approximately 55%.
Unfortunately the new chemical is more expensive so we do not expect to see any
significant financial savings from this aspect.
The additional financial savings will
be achieved through the removal of the requirement to heat a second tank which
will reduce our overall gas cost by approximately 25%. Also the new chemical
can be treated in house thus eliminating the cost of purchasing IBC’s and third
party off site treatment for 2 of the stages giving an additional saving of
over 50% for expense.
With a time frame that has had to be
extended to the annual shutdown a team meeting was called to discuss what else
needed to be done. It was highlighted that some of the stakeholders still had
reservations regarding the project. To help improve the teams understanding of
these issues a Stakeholder Analysis chart was compiled. This gave focus to
which parties still needed convincing of the project and how best to get their
support (see Fig. 4 below).
Fig 4: Stakeholder
Analysis
Control
As the proposed changes to the system
will not take place until the annual factory shutdown the elements detailed in
the Control phase of the project will be based on the planned tests and reviews
that are to be put in place once the changeover has been completed.
From a scheduled team meeting for the
Control Phase it was decided that the SOP’s and Work Instructions should be
reviewed and draft new revisions prepared so that the line operators can be
trained to the new system. These documents have now been created and will be
tested by monitoring and Internal Auditing once the new system is installed and
running to ensure they correctly reflect the procedures being carried out.
The paint chips that were pre-treated
during the factory visit to the firm already using the proposed new chemical were
painted on our automatic line by hanging them on the conveyor after the
pre-treatment stage. Several steel and aluminium chips were prepared and
catalogued. One set of chips (1 steel and 1 aluminium) were sent to our
chemical supplier for them to carry out the NSS testing in their chamber and
another set was sent to our paint supplier for them to carry out the NSS
testing in their chamber. We have asked for the chips to be tested to 1000hrs
which is longer than our customers require and having two test done
simultaneously will allow for a comparison of the results. 1000hrs of testing
equates to approximately 6 weeks so I will probably not have the final results
for the completed report but I have asked for regular interim reports of the
progress and any results received will be included.
As quality of the finish in both
appearance and adhesion are our customers primary requirements a selection of
the remaining chips prepared in the new chemical were subjected to the
destructive tests we are required to carry out. These tests include a
crosshatch test, an impact test, a hardness test and a mandrel bend test. The
results of these are shown below in Fig 5. We also have a digital coating
thickness gauge that we check the thickness of applied paint to ensure that sufficient
paint is being applied and is adhering to the surface.
Fig 5: Destructive Test
Results