|
|
|
|
issue: August 2004 APPLIANCE Magazine
Boiler Production
Simulating Success |
 Printable format
 Email this Article
 Search
 |
|
English boiler producer Baxi Potterton uses discrete simulation software to prevent bottlenecks and improve production line performance. |
|
About a year ago, boiler manufacturer Baxi Potterton Limited of Preston,
England was preparing to place an order for a production line to support
a new type of boiler at its plant in Padiham, Lancashire, UK. This production
line would have numerous automated stations, a human-machine interface
unit, and a barcode data capture function, which was more sophisticated
than what the manufacturer had previously used. The plant that Baxi ordered
it for usually manufactured standard, combination, and condensing boilers,
all of which have different build and testing times than the new kind
of boiler it would now produce.
The group engineer manager, Paul Bilbie, was concerned about the new production
line’s system performance because Baxi had previously experienced
bottlenecks in two of its other production facilities, causing production
losses. Also, after receiving the design for the new production line from
its conveyor supplier, he didn’t think it would run smoothly using
the route the supplier recommended. This introduced a challenge—he
needed to figure out a way to install it, knowing where the bottlenecks
would occur before the production even started.
|
| |
 |
A
3D view of the packaging portion of Baxi Potterton's boiler
plant in Padiham, Lancashire, UK. |
|
|
| |
|
|
Mr. Bilbie, who is responsible for all of
Baxi’s production facilities, maintenance, production engineers,
industrial engineers, and project engineers, started to research discrete
event simulation. After looking into three different companies, Mr. Bilbie
chose to use a simulation software called Quest from Delmia Ltd., the
UK subsidiary of Delmia Corp. (Auburn Hills, MI, U.S.).
Quest is a standard, 3D, digital factory software that is used for process
flow simulation and analysis. It uses a discrete event simulation environment
and pictures of geometric objects from a resource library that enable
users to create a simulation model of a production plant using their plant
layout, alternate scheduling scenarios, and integrated product teams,
as well as items such as buffers, machines, material handling systems,
and docks. Quest incorporates production variables such as physical lengths,
speeds, and a plant layout to analyze the effects on material handling
of equipment and labor.
Mr. Bilbie chose Delmia’s software mainly because it is 3D and it
enabled the boiler manufacturer to use and manipulate software simulation
models it already had. For example, Baxi was able to extrapolate information
from Altercat, a 2D software the company had previously been using, and
combine it with Quest to make it 3D. He evaluated Delmia’s software
by simulating Baxi’s three existing production lines and the new
line proposed by the supplier.
Quest allowed Baxi to develop and watch its production line improve throughout
the design process. Delmia says its clients reduce errors by minimizing
problems and figuring out unplanned costs. With Quest, users can build
a simulation model to the level of detail they want and add more detail
as they go along to keep improving their design.
“Delmia took on the Baxi Potterton project because it was a ‘classic’
manufacturing problem that could be quickly addressed using Quest and
would ensure accurate information upon which Baxi could make informed
decisions,” says Tracy Doran, Delmia’s professional services
manager. “Delmia’s overall goal when working with them was
to demonstrate the flaws with their intended manufacturing concept and
assist them in designing and implementing an optimized line.”
The two companies started by running a 9-day program, using Quest to build
a simulation model. Baxi used its original design (from Altercat), which
gave it a throughput and highlighted where various bottlenecks would ensue.
Then Baxi undertook various tests using “what if” situations
to demonstrate how it could make design changes and still meet the target
throughputs for the line.
According to the companies, this process was complex because the production
lines were based on program logic control (PLC) computers, and there were
a lot of bar code readers and data capturers within the production plant.
Delmia built a production output line into the simulation model, and then
built that line into the 3D program. After first running the program and
seeing all the problems that would have occurred, Delmia further developed
it and added logic control to optimize the lines. The software company
was then able to manipulate the simulation model to find other areas of
concern. When Delmia found a bottleneck when packaging a product, for
instance, it enabled Delmia to modify the packaging area.
“As soon as we built the product, we could tell that the line was
starting to run a lot better and a lot sweeter than it was before,”
says Mr. Bilbie. If Baxi had built the production line without the simulation,
Mr. Bilbie claims the company would have been adding to the problems it
already had on the line, such as carrier starvation. He says he never
knew whether or not the production line received the exact amount of carriers
it required because they travel around the plant automatically. With Quest,
Baxi was able to use the simulation to find that out. What Baxi realized
was that as it put more carriers into the production line, the line started
forming bottlenecks again, which produced gridlock. When it tried reducing
the number of carriers, it starved the line. After experimenting, the
manufacturer saw that 20 to 22 carriers worked perfectly.
 |
Delmia's
Quest software simulation of Baxi's employees at various working
stations within the boiler plant. |
|
|
When the simulation model was built and the results were shown
to Mr. Bilbie, he was pleased. “I had my concerns over the
2 years that I’ve been working with this production line,
and my concerns came to light very quickly when we started to see
the presentation,” he says. In fact, Mr. Bilbie says that
if Baxi had not used Quest, the production line would have been
completely starved of carriers by 2 p.m. on the first day of production,
resulting in a 19-percent loss in production. “It would’ve
been very embarrassing for all concerned,” he says. This would
have induced what Mr. Bilbie calls “total gridlock”
because all the carriers would have been full of partially completed
boilers, so there would have been no way of releasing the completed
boilers from the production line. “We wouldn’t have
been able to produce anything,” he notes.
After receiving the solution from Delmia, Mr. Bilbie showed Baxi’s
conveyer supplier the animated production line in the simulation.
“They were very shocked to see the results and what would’ve
happened if we would’ve gone the way that they were suggesting,”
he says.
Baxi has future plans to build more simulation models with Delmia.
In another recent project, Baxi put together a simulation model
for the front end of its factory, where it is adding four more production
lines. This next generation of lines will be added to all of the
existing lines, and Mr. Bilbie plans to eventually have a complete
3D simulation of the factory running on a computer so he can see
every aspect of production and how many operators he needs per line.
Baxi is also working with Delmia on a Flexible Machining Systems
(FMS) model for machining centers. FMS is a computer and robot-controlled
system.
According to Mr. Bilbie, a critical part of Baxi’s ongoing
success with Delmia is that he can take advantage of the support
days Delmia provides when he needs help with the simulation software.
He suggests that users take Delmia’s advice and understand
that it takes time to build a simulation model. Also, before buying
discrete simulation software, users should have to have a clear
idea of what they intend to do with the software once they have
it. “I knew exactly what I wanted to do with it,” Mr.
Bilbie says. “Delmia helped us to achieve our goals by providing
us with support on building the models.” |
|
|
|
|
|
