Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

EXTRACT LOSS ANALYSIS IN THE

BOTTLING LINE

IROSHA KARUNARAHTNE
Monash University BEng (CHEM) Undergraduate

Table of contents
Introduction
Problem Statement
Process map
Approach
Past data review
Measurement phase
Analyze phase
Implementation phase
Conclusion

Extract loss reduction in in the filling area


Introduction
Beverage manufacturing has been done for centuries. With the revolution of the commercialize
industries the global market has been a real competition. But with the market volatility, the
manufacturers tends to give a wide attention to their economic stability. Due to that the
companies lean towards to improve the processes and compete among others. Financial aspect
of a manufacturing company is always threatened by its product losses. Obtaining the
maximum possible yield for an industrial process is a big hurdle, and it always processed with a
greater challenge. Waste or product losses have been a major burden in production where it
reduces the efficiency and the productivity of the process. Therefore in the recent past waste
reduction have been a dominant force within manufacturing organizations with the growing
economy. Hence the industrial experts are constantly reviewing the performance of their
progressions.
In a breweries point of view extract loss or the beer loss has been the major concern. From the
malt milling to beer filling will include various losses throughout the course. Out of that extract
loss in a bottling area can be a huge financial cost. Since it is the final and the cleanest product
that the customer required. Therefore many beverage companies take an extra effort in cutting
down the losses in a filling line. It has been so clear that the bottle filling needs a distinct
improvement in its operations.

Problem Statement
When it comes to beer bottling there can be several losses and defects. These are some of the
ways that losses can be caused,

Mechanical defects within the equipment.


Lack of technology where it necessary.
Due to procedural issues.
Lack of knowledge and experience within the operators.
Poor material usage and bottle handling.

The main problem that was focused in this context is the extract loss generated in the bottling
area. The impact from the product losses in the filling line have contributed negatively on the
final productivity. Therefore it has a significant effect on the total production output. The
bottling area has the process shown below.

Figure 1: Filling line process map


3

The extract loss points are shown by


1.
2.
3.
4.
5.
6.
7.

in the process map. Those are as following

Volume measurement in the Bright Beer Tank (BBT).


Filler and the crowner
2nd sighter
Pasteurizer
Labeler machine
3rd sighter
Packer

There were several behaviors that how loss occurs in the bottling line, such as breakages, under
fills, overfills, bottle rejects due to bad state, non-crowner bottles, etc. Meanwhile along with
above matters there are some inaccuracies in the volume measurements taken in the BBTs. So
with the limited time a small study was achieved to recognize the main issues that caused the
losses.

Approach
Waste is a financial loss to a company and certainly reduces its profits. With all these in mind it
is vital to conduct a study regarding the losses to find a key solution. As mentioned in the
problem statement the main task was to distinguish between various forms of wastes that have
been generated in the past. Hence the time was utilized more into a cumulative analysis to see
the criticality of the main problem according to the past references. Therefore cost effective
methods are needed to be discovered and implement to eliminate the root causes. The study is
done at Asia Pacific Brewery Lanka in its bottling line. The study was conduct within 2 weeks of
short period.
Nevertheless a small study was performed with the limited time possessed. Therefore the
approach was done by focusing more on to a critical problem based on the historical data. As
the initial step the past data was taken in to account. A brief evaluation was done to foresee
the main problem that contributes to the extract loss. Due to the time constrain as mentioned
above the study was executed by targeting the critical issues.
Next the measuring and data collection phase was performed to collect the necessary data to
emphasis on the key problem. The data was collected and a cumulative analysis was prepared
in order to understand the correlation with the extract loss. Then the problem was reviewed to
see the possible reason and factors affecting it.
The volume measurement in the BBT area was also given a special concern since it also gives an
unknown extract loss to the final production. The main concern over there was due to the
inaccuracy of the volume readings that is taken by the operators. The possible reasons and
improvements have been suggested since not much time was left.
4

Past data review


The past data was taken from the year 2014 January to December. The extract loss percentage
was compared with the KPI value of year 2014 which is 2%.

Actual Extract Loss vs Target Extract Loss (KPI)


4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
Jan

Feb

March

April

May

June

Actual loss

July

Aug

Sep

Oct

Nov

Dec

Target loss

Figure 2: Comparison of the actual extract loss with the KPI value

The above figure shows that the extract loss is above the target KPI value. The average extract
loss is about 2.6% which is well above the target value. So improvements are needed to be
done in order to cut down the loss. Therefore the main losses were analyzed through the data
obtained from the glass breakages and rejection report (Includes in the appendix). It mainly
consist of glass breakages and under fills/rejects categories. The main losses that are in the
report is

Filler crowner breakages.


2nd Sighter under fills/rejects.
Pasteurizer breakages.
Labeler breakages
3rd Sighter under fills/rejects.
Breakages at the packer.

From the above categories the main focus was to analyze the data at 2nd Sighter and the 3rd
Sighter. The main reason was that those losses consist a higher percentage from the total
losses. The following table shows that under fills and reject at the 2 nd and 3rd Sighter consist of
1.71% average loss which is 82.6% of the total loss.

Table 1: Bottle loss parentage based on the respective machine bottle count

Table 2: Bottle loss percentage based on the total production in terms of bottles

The above table shows the total loss through under fills and rejects from the total production.
And it shows that it comprise of 1.76% loss. So the data suggest that the most amount of loss is
generated through the 2nd Sighter and the 3rd Sighter. The total bottle production was obtained
from the monthly production report. The Pareto analysis shown below illustrates that 79% of
the losses are caused at the 2nd and the 3rd sighter. Therefore the measurement phase was
done on those two points to distinguish the impact between under fills and rejects.
6

Pareto analysis on loss points in the filling line


60.00

96.50

100.00

91.64

% loss from the total loss

50.00

100.00
90.00

79.06

80.00
70.00

40.00

60.00
30.00

50.00

48.93

40.00
20.00

30.00
20.00

10.00

10.00
0.00

0.00
2nd Sighter Loss % 3rd Sighter Loss % Pasteurizer Loss % Filler breakage Loss
%

Packer Loss %

Loss Points

Figure 3: Pareto analysis on loss points in the filling line

Measurement Phase
The measurement phase includes data collection and the data analyzing. Initially the data
collection plan was prepared. Due to the presented time and the planned overhaul that started
on 18/02/2015 the data was collected on 4 days for 29 hours. The data was collected on 11th,
12th, 13th, and 16th of February 2015. According to the past data analyze measuring phase done
on two points. The two measuring points were as following
1. 2nd Sighter after the filler.
There were three types of bottles that was put out as waste at the 2nd sighter. They were under
fill low, under fill normal and rejects. The average volume of each type was found using a
measuring cylinder. For each type 20 bottles were used to measure the mean volume. But the
rejects were considered as a full bottle with a volume of 625ml.

Under fills low 314ml


Under fills normal 555ml
Rejects 625ml

2. 3rd Sighter after the labeler.


At this point also there were three types of bottles were identified as waste. They were under
fills normal, rejects and non-crowner bottles. Same as at the 2nd sighter the mean volume was
measured for under fill normal. Over here rejects and non-crowner bottles were assumed to
have the full volume.

Under fills normal 588ml


Rejects 625ml
Non-crowner 625ml

On the first three days the data was collected hourly and the waste was differentiate
accordingly as above at the two points. But on the fourth day the total waste was differentiate
from 6am to 5pm without an hourly count. That was mainly done to identify any odds
throughout the day. The data was measured based on the number of bottles that were put out
of the each sighter. Afterwards it was converted into the extract loss based on the volume by
using the mean values presented above for each type. The data collected on each day is
presented below with remarks for any special incidents.

Table 3: Data collected on 11/02/2015 and 12/02/2015

11/02/2015 from 10am to 5pm


No of
Extract
percentage
Cause
Bottles
loss
loss
Under fills
189
83.27
47.71
Very low
97
30.46
17.45
(314ml)
Normal
39
21.65
12.40
(555ml)
Normal
53
31.16
17.86
(588ml)
Rejects
123
76.88
44.05
(625ml)
No crowners
23
14.38
8.24
(625ml)
Total
335
174.52

12/02/2015 from 10am to 3pm


No of
Extract
percentage
Cause
Bottles
loss
loss
Under fills
180
72.45
47.92
Very low
(314ml)
118
37.05
24.50
Normal
(555ml)
32
17.76
11.75
Normal
(588ml)
30
17.64
11.67
Rejects
(625ml)
114
71.25
47.12
No crowners
(625ml)
12
7.50
4.96
Total
306
151.20

Table 4: Data collected on 13/02/2015 and 16/02/2015

13/02/2015 from 10am to 4pm


No of
Extract
Causes
percentage loss
Bottles
loss
Under fills
143
62.88
50.91
Very low
74
23.24
18.81
(314ml)
Normal
28
15.54
12.58
(555ml)
Normal
41
24.11
19.52
(588ml)
Rejects
89
55.63
45.04
(625ml)
No crowners
8
5.00
4.05
(625ml)
Total
240
123.51

16/02/2015 from 6am to 5pm extract loss


No of
Extract
Cause
Bottles
loss
percentage loss
Under fills
Very low
(314ml)
Normal
(555ml)
Normal
(588ml)
Rejects
(625ml)
No crowners
(625ml)
Total

528

225.61

52.70

288

90.43

21.12

180

99.90

23.34

60

35.28

8.24

310

193.75

45.26

14
852

8.75
428.11

2.04

On 16th there was a special incident regarding reject bottle count. The situation raised by
getting bad condition empty bottles in to the filling area. Resulting in an immediate high
reject count.
Observations made during the measuring phase.

During the hourly counting there were some unusual numbers on rejects at certain
hours.
Most of the time whenever the filler stop operating due to a stoppage the under fill
count increases drastically.

Figure 4: 2nd sighter after the bottle filling


9

Table 5: Difference between the volume loss and the extract loss
Date

Volume
used (HL)

Production
Volume (HL)

Volume
loss (HL)

Volume
loss %

11/2/2015
12/2/2015
13/2/2015
16/2/2015

150.94
107.59
110.96
224.78

148.88
104.16
108.39
217.85

2.06
3.43
2.57
6.93

1.36
3.19
2.32
3.08

Total
extract loss
(L)
174.52
151.20
123.51
428.11

Total extract
loss (HL)

Loss %

Diff

1.75
1.51
1.24
4.28

1.16
1.41
1.11
1.90

0.21
1.78
1.20
1.18

The above data for volume used and production volume was taken as an average for the
stipulated time from the daily production report. The difference between the actual volume
loss and the total extract from the two loss points are shown. Further analyzing is done under
the analyze phase.

Analyze Phase
By using the data that was collected at the measurement phase the following analyze was
prepared. A Pareto analysis was prepared to understand the data that was gathered in the four
days. Two graphs were prepared as below.
1. Analysis based on the percentage extract loss.

Pareto analysis based on the percentage extract loss


50.00
95.94

100.00

45.00

90.00

40.00

80.00
75.29

35.00

70.00

30.00

60.00

25.00
20.00

100.00

50.00
45.31

40.00

15.00

30.00

10.00

20.00

5.00

10.00

0.00

0.00
Rejects

Underfills Normal

Underfills Low

No crowner

Percentage loss
Cumulative loss
Figure 5: Pareto analysis based on the percentage
extract loss

10

From the above graph it clearly shows that rejects and under fills have an equal impact on
the waste that is generated at that loss point in terms of the extract. There is 45.31% of
rejects and 50.63% of total under fills in this category based on the total extract loss. So it
suggest that an outside factor is present with the extract loss in the bottling area due to the
rejects. The extract loss from rejects at this point is a major worry that has been neglected
completely.

2. Analysis based on the percentage frequency.

Pareto analysis based on the percentage frequency


45.00

96.18

100.00

100.00
90.00

40.00

80.00

81.25

35.00

70.00

30.00

60.00

25.00

50.00
20.00

42.60

40.00

15.00

30.00

10.00

20.00

5.00

10.00

0.00

0.00
Rejects

Underfills low
Frequency %

Underfills Normal

No crowner

Cumulative %

Figure 6: Pareto analysis based on the percentage frequency in term of number of bottles

The above analysis was based on the frequency which is the amount of bottles put out from
the two sighters. From there also it proves that rejects bottles have a huge influence on the
under fill/ reject category. From this figure it suggest that 53.58% of the bottle includes
under fills and 42.6% are rejects.
So from the both Pareto analysis it suggest that rejects have a dominating effect on this
waste category. The data clearly interprets that an external issue was causing extract losses
in the filling line which is a major worry. The two main problems that are present at the
moment are under fills and the rejects.
11

The under fills can occur due to technological and mechanical concerns. These issues are
hard to analyze during a production time and with a short period of time in hand. When it
comes to technology its always difficult to predict its faults. But definitely it can have
certain mechanical complications which is need to be found through a deep analysis.
According to the machine manual it suggest some of the following reasons for causing
under fills.

Filler bowl level too low.


Liquid cone defectives.
Vacuum and snifting valves have got stuck or leaking.
Contact element of the bottle lifting element is too low
Operating butterfly not correctly adjusted.
Air tube bent.
Filling valve not operating properly
Contact pressure of bottle lifting element is too low.
Air tube deflector defectives.

During the data collection phase under fills were caused due to the low filler bowl level. It
mainly occurs during filler stoppages. Once the filler stops due to a line stoppage or for any
other reason and once it starts it tends to cause under fills. The main reason is during a
minor stoppage if the filler bowl level is low it wont fill automatically. So once the
production starts it will cause under fills. And another way it happened was due to excessive
foaming. Some of the bottles that were put out of the sighters contained heavy foam which
had little product in it. But completely to reach for a full and final solution the filler needs to
be investigate thoroughly for long period of time.
Meanwhile more than focusing on a mechanical fault it was much simpler to look at the
rejects which are caused externally. In the other hand reject bottles plays a serious threat to
the amount of waste produced. Bottles are put out as rejects depending on its condition
and state. The following are some of the reasons for bottle rejects that were found at waste
points.

Damaged bottles.
Small pieces of dirt inside the bottles
Marks inside the bottle wall.
Labels are stuck inside the bottles.

Even from the results it suggest the criticality of the problem raised by rejects. A reject bottle is
considered as a full bottle which is a huge loss. Therefore in under fills also it can consist of
rejects which is neglected in this study. But the main point that is interpreted by the result is
amount of loss that cause due to rejects. These are some of the reasons why reject bottles are
found as waste with product in it.
12

The operator at the empty bottle sorting sighter misses the bad state bottles.
Washer faults or defects.
The operator at the 1st sighter cannot identify clearly regarding the bottle condition.
The operator at the 2nd sighter cannot identify clearly regarding the filled bottle
condition.
Accepting bottles that are in real bad condition.

Another problem that was identified during the study was that the extract loss obtained from
the loss points has a difference with the actual extract loss in terms of volume. One of the main
reason for this problem is caused due to inaccuracy in the volume measurement. From the
table 5 it suggest that there is a significant difference between the two values. The extract loss
only includes the under fills and reject category. But according to past data it suggest that
breakages only has a small effect on the total loss. As presented in the past data review the
breakages only includes 0.45% of the total bottle production in terms of number of bottles
produced. And its only 21.7% of the total loss according to the Pareto analysis as shown in the
Figure 3

The problem with the volume measurement is due to the inaccuracy in the volume
measurement through the volume indicator shown on the bright beer tank. With its shape it
does not have a constant volume horizontally. The indicator is shown below.

Figure 7: Volume indicator on the BBT


(Tank 704)

Figure 8: Volume indicator on the


BBT (Tank 701 to 703)
13

The volume is measured through the indicator by an operator. Since these tanks very tall the
volume measurement at the top level is hard. So the measurement that are obtained at that
level has certain errors. So with the human error and the indicator error will definitely give an
inaccurate final volume. A volume measurement between two values or markers shown in the
indicator is rounded to the nearest value. And the value is converted from the volume table
provided from the tank manufacturing company.

As an example according to explained above, if the


magnet indicates a value between 1150 and 1100 it will
be rounded to the nearest value depending on the
magnetic indicator. Generally in tank 701 to 703 includes
1.5HL between two markers (values). In tank 704 it
includes 3HL approximately. But that value is only an
average value due to the tank shape. So in that way it
contributes to the extract loss percentage at the end.
Even though its seems to be a small thing through this,
the data from table 5 shows a considerable difference
between the extract loss and the total volume loss.

Figure 9: Close look on the volume


markers on the volume indicator

14

Implementation Phase
Due to the limited amount of time it was unable to implement ideas or solutions for the above
mentioned problems and for its causes. But however it was able to go through some of the
current implementations.
In order to reduce and prevent the amount of reject bottles that are coming in to the filling
area following One Point Learning (OPL) board is displayed. Therefore the operator can clearly
identify the type of bottles that are needed to be reject.

Figure 10: One Point Learning (OPL) displayed in the bottling


line

15

Suggestions

Operators needed to be more aware of the fact that reject bottles causes a huge extract
loss. Therefore the necessary knowledge regarding rejects must be given to the
operators.
Operators can inform the supervisors regarding any unusual amount of waste that is
been put out of the sighters. Therefore the supervisors can looked into that
immediately.
Currently displayed KPI boards can be filled properly on a daily basis review the
performance. Though the boards are been displayed nothing mentioned in the boards.
Hence the operator supervisors and the respective management can review into any
unusual data.

Figure 11: KPI Board in the bottling area

16

Figure 12: Bottling Performance board in the bottling area

During a shift change over the specific performance to the shift can be discussed briefly
among the supervisors and the line leaders. If there is any unusual amount of loss in the
shift it can be displayed on the respective boards daily. Therefore the performance can
be evaluated weekly and discuss it in order to improve it.
Standard operating procedures can be given to the operators. And put more OPLs
around the bottling line regarding the small things that can be very useful.
For the exact volume measurement the flow meter can be replaced before the filler.
The figure below shows the current flow meter in place which is not working.

17

Flow meters are generally very accurate. Therefore the volume that is been used for production
can be accurately measured. So it will eliminate the extract loss that is been caused due to error
in the volume reading.

Conclusion
The bottling process of beer has faced many obstacles which could be alleviated through a
focused process improvement. A Pareto analysis can be used as a very effective tool to identify
key areas in the process that contributes heavily to cause product losses. Therefore it could be
benefit from a focus improvement initiative, thereby benefiting the overall company. An extra
effort needed to be put to eliminate the waste problem such as rejects. This study gives a
certain insight regarding the waste influenced by under fills and rejects. It gives the highest
amount of waste in the bottling line. For under fills the root cause may be due to its
technological and mechanical defects. But certainly rejects are mostly caused due to an outside
factor.
With the short time that was available it was unable to implement any recommended solutions
for the area that were identified to have a negative influence on the production process. It is
crucial to always ensure that the workers comply with the standard operating procedures in
order for waste to be reduced.

18

You might also like