7
$\begingroup$

I understand that the concept of line balancing is such that each station takes exactly the same amount of time to process one unit. Clearly, this is an idealization and real manufacturing plants will have bottleneck stations and underutilized stations.

I'm wondering if someone has an idea on what numbers are typical in various industries? For instance, if there is a 10% take difference between the slowest and fastest stations, is the line considered balanced enough?

$\endgroup$
  • 1
    $\begingroup$ Queuing theory is a pretty well understood branch of statistics. Cost vs. schedule is a standard input to these models. $\endgroup$ – Carl Witthoft May 4 '16 at 11:29
9
$\begingroup$

I can answer for the Chemicals Sector I'm familiar with: Overall line balancing is quite poor between processing operations. 10% difference would be surprisingly good. 50% difference isn't uncommon.

Several reasons for this I can think of:

  1. Lot of equipment comes in discrete sizes. e.g. Compressors, Centrifuges etc So often the next size higher means you have a lot of slack built in that equipment from Day-0.
  2. Often the exact max capacity of a complex piece of equipment like a distillation column is not known with 100% accuracy. So designers incorporate a factor of safety in mentioning the rated capacity (that will become a contractual obligation). Once plant is operational it can be pushed to max capacity which is often higher than the on-paper capacity the equipment was designed for. Ergo now other pieces of equipment become bottlenecks. And you have an unbalanced line.
  3. Some equipment like Heat Exchangers or Plug Flow Reactors will foul over their working life and hence become less effective. There can be a wide margin between their Day-0 throughput and 6 months down the line. As a result, other equipment not suffering from such in-service derating will end up showing excess capacity.
  4. Often Capex pricing of specific equipment is highly non-linear with respect to capacity. e.g. You could be adding 50% extra capacity at only 10% extra price. So designers end up having unbalanced lines simply by hedging their risks by over-specifing where the cost penalty isn't high.
  5. Over the long life of a plant (e.g. 20 yrs) some innovations come up where with modest investment a equipment can be significantly boosted in throughput. e.g. Better generations of catalysts. The parts of the train that cannot be innovated upon then become bottlenecks.
  6. The lead times for new equipment commissioning can be long. e.g. 3 years. So sometimes management will knowingly add larger equipment, consciously realizing that the new equipment will stay underutilized for a while till the other parts of the line can be upgraded.
  7. Employees & site-engineers come up with operating innovations that debottleneck a specific equipment. Suddenly the rest of the line is underutilized.

    8.Chemical processing lines can make multiple products. e.g.

    A-->B-->C-->D-->E

where both C & E are salable products. The plant was designed line-balanced assuming you'd sell 500 tons of E & 300 tons of C every month. Unfortunately 5 years down the line the market for C is bad. So parts of the line pre-C become underutilized although the rest of the line is still balanced.

| improve this answer | |
$\endgroup$
2
$\begingroup$

It will depend on the industry. Competition drives new processes, techniques and technology that improve efficiency. Also, the larger the market, and market share of a company, the greater the economies of scale. Also, the certainty of a particular station being able to process parts at a precise speed will depend on factors such as its complexity and the nature of the parts or materials used to make it, changing perhaps from one hour to the next, or one day to the next, for example, between 80% and theoretically at least, 100%.

For example, a lot of laboratory glassware is surprisingly still largely hand-made. Tools and techniques have improved, and no doubt will continue doing so. But automation always becoming cheaper. But the bespoke nature means it is unlikely to ever warrant factory lines.

But on the other side of the scale, although I don't have figures, common hardware such as nuts and bolts are produced on such large scales, and the materials so consistent, and the process so completely automated, that the utilisation of posts is with little doubt above 90%. But this is then open to debate exactly what the definition here of efficiency is. For example, if it is the capacity of an average person with a given skill set, or a given machine at full speed, the economy of a machine at a certain speed: full speed may incur higher energy consumption per part, dramatically increase wear on its parts, or reduce quality of parts produced.

But how far do companies go? It's down to their bottom line; whatever makes more money.

| improve this answer | |
$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.