Making time for profits

Newton Industrial Consultants specialises in identifying small but powerful changes to the production process that can improve plant efficiency and raise productivity by at least 10%

Whether you are looking to increase throughput to meet demand, reduce process hours, or increase capacity to take on a new contract, as a manufacturer of bulk pharmaceuticals, batch cycle times usually offer the biggest lever to improvement. Reducing them rarely requires capital expenditure and can give huge rewards.Achieving a significant reduction will not be easy. If it were, it would already have been done. Only by challenging every part of the cycle and driving through the necessary changes, will anything be achieved. It is usually best to begin with the non-critical parts of the process, such as filling and emptying, as they are usually less contentious. Getting agreement around a few quick wins will be critical to getting the belief and commitment behind batch cycle time reduction.

extra capacity

Reducing batch cycle times offers the opportunity to produce more product in the same period of time. What you do with the extra capacity - take on a new contract, fulfil more orders or simply reduce process hours - determines what it is worth. If the extra capacity can be sold, then with no extra costs other than the raw materials, energy and shipping, the extra production has a very high profit margin - commonly 3-5 times the usual margin. For example, if raw materials, energy and shipping were 50% of the sales price for a company with £100m sales, which currently makes £10m profit, then a 20% increase in output and sales would be worth 50% of £20m in extra profit, creating a doubling of profits.

Understanding the value of having the extra capacity makes it easy to calculate the value of a cycle time reduction on the process. At one UK plant, limited by a batch process, each second saved on a 42-minute cycle was worth £2,500 a year, or £150,000 for each minute. It is worth doing the calculation for any such project as the value might be surprising and will drive everyone to challenge each minute or second of the cycle time.

Before we can begin reducing cycle times, we need to understand what makes up the cycle time and how much it varies. The first step is to take a sample of batch runs - perhaps 10 or 20 - and break them down into the individual steps. For each step examine the average time and the best time: if the two are very different, this suggests large potential for time reduction. This information is more easily reviewed if presented as a gantt chart (figure 1). The green portion of each step is the best-observed time, and the red portion is the difference between the average time and the best time. If we could always do our best time, in this case it would mean a 23% cut in cycle time.

There are now two areas to look at to reduce the cycle time:

solving problems to eliminate the differences so that we can always achieve the best time or at least get closer to it

reducing our best time

To achieve the biggest impact on the cycle time in the shortest period, we need to know which problems have the highest value and solve those; it may also be possible to identify some quick wins. The first step is to establish and compare the values of different problems to assess how much resource or effort to focus on each. To do this, we can form two paretos, one for 'red' time and one for 'green' time, prioritising the issues in order of size (figures 2 and 3).

Since we do, at least on occasions achieve our best, then we would expect that the problems behind the 'red' time are solvable, so the whole 'red' value is an achievable potential. Clearly we cannot regard the green time in the same way, but we would expect to be able to make an impact on it. Certainly, a 10% impact is usually achievable. If we then regard 10% of the 'green' time as achievable potential, we can combine the paretos (figure 4). This gives a guide to which areas are likely to yield the most cycle time reduction and will enable decisions to be made about where and how much effort, time and resource should be expended.

At this stage, any quick and easy wins may be identified. Sometimes by exposing the size of different areas of potential, one emerges which either was not known to be a problem, or was not thought to be of any value. For example, one UK plant found that there was a step in the process, which had historically required a wait while an operator carried out a manual check on salinity, but this step was not removed when a new salinity sensing system was added, and so caused a delay if operators did not step the process on quickly.

As well as bringing real results, quick wins also encourage the cycle time cutting initiative to gain momentum by breeding belief that improvement is possible and encouraging further efforts.

largest potential

The next task is to reduce the steps with the largest potential. The general approach is to understand what things determine the time taken to reach the end of the step and then use those 'levers' to reduce it. For example, on a step involving heating to a set point, these levers would be the time heating starts, the heat added, the heat lost and the mass of material to be heated.

On a mixing cycle you might look at the time mixing starts, the speed of the mixing blades, and the size and shape of the mixing blades. On an unloading cycle it might be the time unloading commences, the discharge pressure, the material viscosity and the orifice size. By challenging each of the levers in turn, it is possible to have significant impact on the higher-level issue: the time the step finishes.

A simple example is a US factory, where a manual check on the process was required part way through the batch cycle before it could be stepped on. The time it takes is determined by the levers: how long it takes to notice the event, how long it takes to carry out the check, and how long it takes to step the process on. This relied on the operator noticing the event from a small flag on the screen, going out onto the process floor to perform the check and then returning to the screen to step it on. The check itself was executed quickly and efficiently, but the time to notice the event and the time walking to and from the process floor offered room for improvement.

By changing the warning to a full flashing screen, it was very quickly noticed, and by adding a button on the process floor which meant the process could be stepped on without returning to the screen, the company saved an average of 12 minutes in a three-hour cycle: a 7.1% increase in capacity.

significant savings

A more complex example is a heating cycle in a heat treatment process. The process operates in a vacuum, and the heating began once all the material for heat treatment was added and the vacuum achieved. The levers on the time this step finishes are: time heating starts, heat transferred to the material, heat lost from the material and the mass of material. The first thing to be challenged was the time heating began. By starting heating when the vessel was half full, the company saved 18 minutes heating time on the 14-hour cycle.

The other levers were then examined. Heat loss from the vessel (and hence the material) was found to be very small, and changing the mass of the material would mean a reduction in batch size, which would reduce overall output. This left only heat transferred to the material as a source for improvement. Under a vacuum, radiation is the only significant form of heat transfer. By switching the vacuum on only when the material had reached 95% of temperature, conduction and convection through the air could also provide heat transfer and an additional 17 minutes were saved from the cycle time. The same approach can be applied to any step in any process.

It is important always to challenge the process and ask: How much further can we reduce this? And what do we need to do now to reduce it further?

major variation

In the unloading step of a UK chemical blending process, there was significant variation between different lines. Unloading occurred through gravity pushing the dry chemical through an orifice. The orifices, although the same diameter, had different sieves under them. By changing them all to the larger sieve size, an average of 5 minutes was taken from the 2-hour cycle.

But further challenging called into question the reason for having a sieve at all. The sieve was there only to stop any loose bolts from the inside of the vessel falling into the blended product. By enlarging the head on these bolts, the sieve could be made three times as coarse, saving an additional 8 minutes.

Often there will be a number of steps that could be performed in parallel with little change to the system, effectively 'hiding' the step from the overall cycle time. A large gain for a small US pharmaceutical firm was to eliminate an average 75-minute wait on an 8-hour batch process. This was in place to prevent damage to the product by overheating if the following process was not available, by heating the product only when the following process was 75% of the way through its cycle. The step was eliminated and on the rare occasion that the following process was unavailable, the product was cooled to a safe temperature while it waited.

Most of these examples were only trivial changes to processes, requiring no capital expenditure, but they yielded big improvements in performance. But a high level of commitment is required to get results and it will be necessary to dedicate resource exclusively to the task. But with a disciplined and rigorous approach reduced cycle time can quickly be turned into increased profits.