Measuring downtime in liquid batching


Thread Starter

Mal Steel

We have a process of 4 liquid brewing vessels in series,(Brewery) and are interested to measure downtime and resons for falling behind production schedules

But its difficult to measure what you can't define
Most 'downtime' models are aimed at Packaging lines and defining downtime reasons are easy
Here the vessels typically hold a brew for 60 -90 minutes (recipe driven) before passing to the next stage, but hold ups or slow brew times in 1 vessel dont always slow down the vessel upstream, depending on where it is in its own schedule.
Typically Vesssel 2 is slowest and the others may wait on it, but a problem in vessel 1 or 3 can impact the whole line if it takes an hour or more to repair

Has anyone every worked on a similar problem or know where I can find data

Best Regards

All process interruptions should be logged. Production lost time if any should be recorded. While your process is a single stream it may appear necessary to record only those events which affect production but if the facility is to expand it is essential to know the true availability of all stages in the process. Similarly if capital is to be spent to overcome a specific delay then it is necessary to know what gain will be achieved in order to calculate the economics of spending the capital. That is, you need to know what the next bottleneck will be if you eliminate that specific delay.

Even in highly automated facilities it is not generally possible to automate the collection of downtime data. It is usually an operator function. It can be just a written log but some systems can be made to identify downtime and prompt for reasons for delays and have operators fill in the data manually. As with any manual process operation it may need a reasonable amount of discussion with management, process operations and maintenance people to understand the reasons and agree the best methods of data collection. If the process is continuous and a
lot of activities occur during the day shifts then consider holding duiscussions on the back shifts if operators have more time available then. They will appreiate your consideration and committment.

There also needs to be an agreed visible follow up process of analysing the data, discussing the causes, determining solutions and taking action to improve the process. Many of the solutions are likely to be changes in maintenance or oprating
practices so again those people need to be involved in determing the best solutions.

Eliminating process variability and de-bottlenecking to achieve more production from a given facility without increased costs in energy, manpower etc are about the best investments that can be made and can be very rewarding for all involved.

Another step you may wish to take is to look at how things are done in other plants particularly those that are bench-mark in their field. We have continuous chemical processes and have gained significant benefits from employing methods of
continuous improvement used by Toyota in car production. The processes may be different but many good management and operating practices work on any process.

Take your problem to John Boland at Visibit.

Walt Boyes

Walt Boyes -- MarketingPractice Consultants
[email protected]
21118 SE 278th Place - Maple Valley, WA 98038
253-709-5046 cell 425-432-8262 home office
fax:801-749-7142 ICQ: 59435534

"Strategic marketing, sales and electronic
business consulting for the small and medium-sized
co-author of "e-Business in Manufacturing: Putting the Internet
to Work in the Industrial Enterprise" ISA Press--September 2001 ISBN:

Chiron Consulting


What you're describing sounds like a batch system where transitions from one stage to the next are driven by recipe/strategy, but dependent on the
availability of resources. Think of the system as an event-driven state machine.

Your batch is sitting in a vessel (batch state A). After awhile, the recipe says it's time to move to the next vessel (transition to State B on
a "timer elapsed" event.). But the next vessel isn't ready, so the batch sits where it is, waiting for a necessary resource. Eventually, the next vessel becomes available ("notification" event) and the batch moves to state B.

If you can characterize your process as a state machine, and track the events and transitions, you can do the analysis you want.

In this case, we can make the state machine more rigorous, by defining a state A', in which the batch is in the first vessel, ready to move, but
waiting for the next vessel.

Now the timer event causes a batch transition from state A to A'. If the next vessel is available, then the notification event is immediate, and the batch moves to state B. If the next vessel is not available, the batch sits in state A' until the vessel becomes available. You can define your state machine such that certain states represent certain kinds of hold-up in the process. Then track how long each batch spends in each state during its lifespan.

We do something similar with the raw material batching system that feeds a collection of smelting furnaces.


Greg Goodman
Chiron Consulting

So the min/max batch time is 60 to 90 minutes. You must be allowing that range because of what happens and when the other batches can or cannot keep up.

The tanks are in series.

A continuous level transducer can be used to prevent beer from leaving one tank or entering the next tank - based on level.

For a low cost way of looking at history - the transducers can operate charts showing level, date and time.

We have a brand new controller that has 4 programmable set point relays, 1 bar graph and a 3 diget display for real time looks at what is going on. (picture not online yet)

Thanks Bob Hogg
Start with a detailed and accurate flow chart of the process.. i have found that just doing that will reveal the best ways to monitor the processes and the conditons that make a 'downtime' event true..
Good Luck..
There is a software called throughput, which i believe works on the principles developed by Goldratt. This software gives identifies constraints and provides proper control over constraints plus there are several additional features.

The software was basically developed for the mechanical industry but can be used for most processes where there are resource constraints.

Software provides detailed reports etc.