10th International Conference on Through-life Engineering Service

16-17 November 2021, University of Twente, The Netherlands

Reducing the cost of tea production through employing usage

based preventive maintenance (Case study in Kenya tea
development agency factory KTDA Chebut)
Kiprotich Evans*, Benard Kipsang
410082, Changsha city, P.R.China
486-30300, Kapsabet, Kenya
*Corresponding author: Tel.: +254715970745; E-mail address: kiprotichevan@gmail.com

Abstract

In a continuous production line, most of the machinery requires maintenance to ensure that they are working properly since
the breakdown in one machine stops the whole production line. Therefore, preventive maintenance is always scheduled to
ensure that machine parts are replaced and maintained. During replacement, production is at halt and thus the production output
is decreased. An effective scheduling of maintenance can help to reduce downtime and reduce the number of schedule
maintenance. This case study investigated how the use of usage-based preventive maintenance in place of periodic preventive
maintenance can help to optimize maintenance further by; reducing maintenance cost, extending lifetime of an equipment and
reducing the overall production cost. To perform the optimization, the t-test hypothesis was formulated upon the reception of
the actual production data. The hypothesis was used to investigate if the actual production is same or different from the design
potential production capacity. The results and findings showed that the actual production was different from the expected
design potential thus the alternative hypothesis is not rejected. The efficiency test was found to be 79% showing that the
maintenance set for the system is under-utilized which caused over-maintenance of equipment. Under-utilization attributed
majorly to the periodic maintenance applied even at low input of tea leaves to the factory. The results showed that optimized
usage-based preventive maintenance can be used to save up to 10 scheduled maintenance per year. Reduction in numbers of
maintenance directly reduce maintenance cost, hence, minimization in overall production cost. From the results, this study
lays a basis to optimize production in any organization by optimization through minimization of input resources and
maximization of outputs. The optimization strategy of maintenance applies to both product and service organizations.

Keywords: CTC black tea processing; Preventive maintenance; Usage-based preventive maintenance; Optimization;
Efficiency; saved maintenances

1. Introduction Optimization of the maintenance process through

usage preventive maintenance can further improve
Preventive maintenance is the work carried out on the production system and impact positively on
equipment in order to avoid its breakdown or economic value of the company.
malfunction. It is a regular routine action done on an The present study aimed at determining whether
equipment in order to prevent its breakdown [1]. or not the cost of tea production is reduced
Product and service production organization have significantly by usage based preventive maintenance
since improved their maintenance and repair skills system. Hence, it demonstrated how usage-based
and pay more attention on maintenance. Previously maintenance can be of value if implemented in the
maintenance was never of much concern, however, factory.
as technology improved, mechanization and The outstanding contributions of this study
automation increased, therefore arose the need for includes; increasing tea produced through reduced
good maintenance services. As technology advances, equipment downtime, reducing repair and
maintenance has also advanced in order to reduce maintenance cost, increasing profit margins of the
downtime and improve safety [2]. company through less downtime of equipment, and
In the 21st century, many producing organizations extending useful lifetime of the CTC (cut, tear, curl)
have embraced new technology and adopted rollers.
preventive maintenance for monitoring their The company works on three shift system per day
production line state. The situation is same for Kenya and on a weekly basis, the company does a routine
Tea Development Agency (KTDA) Chebut tea periodic maintenance normally schedule on
processing factory. The challenge, however, of using Mondays. During this maintenance period the
Preventive maintenance is over or under company process has to stop for at least two shifts.
maintenance if not optimized. This paper quantitatively analyzed the production

Peer-review under responsibility of the Programme Chair of the 10th International Conference on
Through-life Engineering Services.

Electronic copy available at: https://ssrn.com/abstract=3949300

 Evans et al. SSRN-TESConf (2021) 016

and maintenance data of KTDA Chebut factory and Spain, Hernandez-Chover et al found out that the
showed clearly how optimization on existing maintenance costs represent on average 11 % of the
preventive maintenance can be of significance on the total operational costs. In Greece, Tsagarakis et al
economy of the tea factory. The study is focused at compared the maintenance costs of different
optimization of maintenance cost using usage-based activated sludge wastewater treatment technologies
preventive maintenance in place of periodic reporting mean values that range from 6 to 9% of the
maintenance being practice in the company. overall annual Operation Management costs [9].
Preventive maintenance tasks usually come with
2. Literature Review a cost and are very vital and necessary to ensure that
the equipment is operating properly and minimize
Maintenance, repair and overhaul or operations the risks inherent to possible breakages and
(MRO) are normally referred to as the functions of downtime [10]. The absence of optimal preventive
maintenance [3], are what makes up maintenance the maintenance always tends to increase the costs set
very important component of any production. aside for repairs and is evident as many
Historically, maintenance was not a great operation managements report from different kinds of
with great importance in the past when production infrastructure such as bridges, roads, rail networks
was done in small scale. According to Ben Niebel, or water facilities, as they encounter various
author of Engineering Maintenance Management problems [11].
(Industrial Engineering), stated that in 1969 The rising costs allocated for repairs and
maintenance was 1-17% of total workforce [4]. breakdowns [12], may decrease machine
However, as advances occurs in technology more productivity as well as entail environmental and
types of machinery were made available, and the social risks. In this case, therefore, preventive
need for maintenance personnel increased. maintenance, activities planned to maintain the
Though maintenance departments are not directly equipment or machinery in optimal conditions, is
related to production, they can still have a profound needed to reduce unexpected breakages and reduce
effect on the quality and cost of production [5]. the cost of repairs and maintenance [13].
Maintenance is really developing and using it to
reduce downtime, breakages and risk will not be 3. Methodology
enough in the currently competitive economy where
cost effective organization is highly demanded. The In order to achieve the objective of this study, the
organization may be working smoothly because of relevant data was collected from the case study
preventive maintenance but yet the cost of input on company. Data was obtained from daily records of
maintenance can be very high. Optimization of production in log books and company production
preventive maintenance of a system is necessary databases.
now to an existing maintenance strategy in order to This study applied the statistical tools in Minitab
realize more reduction in cost of maintenance and and excel to analyze and check if the machine usage
repairs thus reducing total production cost. limit compared to actual production gave significant
difference which can be relevant to make adjustment
2.1. Cost of maintenance
towards reduction of production cost. Analysis
involved the raw data from production line for daily
Economic aspect has not been of more concern in
production on weekly time period before the routine
technology growth in the past as it is in the current
maintenance.
market and economy.
Some of the important equations used in the
Maintenance cost being a component of
analysis part of this research were; T-test formula
production cost, thus, equals certain percentage of
and efficiency formula.
the total cost and varies depending on level of
optimization and event technology applied. It is
3.1. T-test formula
known that maintenance costs of waste water
treatment plant can reach up to 25 % of total
The t-test is any statistical hypothesis test in
operational costs. For example, in United States,
which the test statistic follows a Student's t-
Biehl and Inman study shows that this percentage
distribution under the null hypothesis.
could represent the 15 % of the total operation cost
The t-test was used to determine if there is a
[6]. In Turkey, waste water study of Turkmenler and
significant difference between the design
Aslan indicates that maintenance costs represent
comfortable capability of production line and the
approximately the 10 % of total operational costs
actual production in the line.
[7]. In some European countries similar values are
found, for instance, in Germany, these costs can 𝐱̄ – 𝛍
range between 15 % and 25 % [8]. Similarly, in T value t= (1)
𝐬 / √𝐧

Electronic copy available at: https://ssrn.com/abstract=3949300

 Evans et al. SSRN-TESConf (2021) 016

Where, Histogram of GREEN TEA IN KG

x̄ = Observed Mean of the Sample 40
Normal
Mean 39651

μ = Theoretical Mean of the Population StDev 4878

N 301

s = Standard Deviation of the Sample 30

n = Sample Size.

Frequency
20
3.2. Efficiency formula
10
Efficiency is the ability to avoid wasting
materials, energy, efforts, money, and time in doing
0
something or in producing a desired result. In a more 28000 32000 36000 40000 44000 48000 52000
GREEN TEA IN KG
general sense, it is the ability to do things well,
successfully, and without waste [14]. Fig.1. Histogram of sample data and normal curve

𝑜𝑢𝑡𝑝𝑢𝑡 Normality test was done by considering the p-

Efficiency = ∗ 100% (2) value of the sample. The p-value of p>0.05 shows
𝑖𝑛𝑝𝑢𝑡
that the data is normal and p<0.05 shows that the
4. Results and Discussion data is not normal. The figure below shows the
results for p value of the green tea leaves data
This section deals with the comprehensive samples obtained from Minitab.
analysis, discussion of the calculated results and
interpretation.
The factory is potentially design to handle
approximately:
 15,000,000kg of green tea leaves annually.
 300,000kg before maintenance
 50,000kg green tea leaves daily.

4.1. summarized data

The table below shows summarized data for year
2020.
Table 1. Monthly green tea leaves data summary
Months Working days Daily average tea in kg
January 24 34833
February 25 31240
March 26 35000 Fig.2. Histogram of sample data and normal curve
April 24 40042
The P-value = 0.054 confirms the data is
May 26 44769
normally distributed.
June 24 43333
July 27 38741
4.3. T-test
August 26 40231
September 26 44654
From equation (1) that is:
October 25 47320
x̄ – μ
November 25 38480 t= (1)
s / √n
December 23 36739
average 39615 We can calculate the variables and substitute for
t from table 1.
4.2. Raw data normality test.
Sample mean = x̄ =39615 kg
Upon the reception of green tea daily processed Population mean= µ = 50000kg
data, normality test for the data was run to check if Deviation=s =4771kg
the data is normal. Sample size = n =12 months
The histogram data points and a normal curve Formulation of the T-test problem
drawn and compared. Null hypothesis H₀: x̄=μ = 50000kg.
Alternative hypothesis H₁: x̄≠ μ ≠ 50000kg.
The,

Electronic copy available at: https://ssrn.com/abstract=3949300

 Evans et al. SSRN-TESConf (2021) 016

11935000𝑘𝑔
Null hypothesis represented the assumption that Green tea daily average =
301𝑑𝑎𝑦𝑠
the actual production had no difference from the
potential of the production line. = 39651kg per day computed using the yearly
While, raw data
Alternative hypothesis represented the Also,
assumption that the actual production was different ∑𝑜𝑓 𝑚𝑜𝑛𝑡ℎ𝑙𝑦 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑔𝑟𝑒𝑒𝑛 𝑡𝑒𝑎
from the potential of the production line. = 39615kg
12
Performing the test at 95% confidence limit =39615kg per day computed from monthly
Degree of freedom n-1 =12-1=11. grouped data.
Solution: Solution
39615
t=
x̄ – μ
=
39615−50000
= -7.54 Efficiency= ∗ 100%
50000
s / √n 4771 / √12
= 79.23% from monthly grouped data
t – Value = -7.54
The absolute t value = 7.54. 60000
Critical t value at 95% confidence interval =2.2. 50000
Therefore, 40000
Absolute t value > critical value 30000
= 7.54>2.2 thus, null hypothesis is rejected. 20000
The rejection of the null hypothesis, shows that 10000
the system of production, is not fully utilizing the 0
potential capability as design. daily potential daily actual
capacity production
4.4. Efficiency
Fig.4. Daily potential vs. actual production
From equation (2).
Generally efficiency can be calculated by: 20000000
𝑜𝑢𝑡𝑝𝑢𝑡
Efficiency= ∗ 100% (2)
𝑖𝑛𝑝𝑢𝑡
15000000
4.4.1. Economic Efficiency 10000000

Economic efficiency is when all goods and 5000000

factors of production in an economy are distributed 0
or allocated to their most valuable uses and waste is annually annually
eliminated or minimized. potential actual
Economic efficiency is basically just a measure capacity production
of how good things are economically, compared to
how good they could potentially be. Fig.5. Daily potential vs. actual production
The formula for determining economic efficiency
is as follows: 11935000
Economic Efficiency= Efficiency= ∗ 100%
15000000
𝑎𝑐𝑡𝑢𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑒𝑐𝑜𝑛𝑜𝑚𝑖𝑐 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒
(3) =79.57% from yearly raw data
𝑝𝑜𝑡𝑒𝑛𝑡𝑖𝑎𝑙 𝑣𝑎𝑙𝑢𝑒 𝑜𝑓 𝑒𝑐𝑜𝑛𝑜𝑚𝑖𝑐 𝑣𝑎𝑟𝑖𝑎𝑏𝑙𝑒

The low efficiency in this case was majorly

Potential economic value design for the line, is
attributed to the natural cause of less tea available
the total value of input resources and services into
during low production seasons, a factor that the
the production line. This economic input represent company could not control. . As shown, the
the economic value of; labor, materials, power, maintenance period was likely to be reached always
maintenance, material handling, storage and other when the system has not exhausted its design
production factors potential and thus, it resulted into low production
The potential economic input was capable and efficiency. Approximately 79% production potential
responsible of handling approximately 50000kg and ability was utilized while 21% was waste.
15000000kg of green tea leaves each day and year The production line according to the design
respectively according to the design specification of specification was designed to handle 50,000kg of
the production line. green tea, daily without exhaustion of the line
From table 1 machinery and the resources required for the
production.

Electronic copy available at: https://ssrn.com/abstract=3949300

 Evans et al. SSRN-TESConf (2021) 016

The design potential capacity was supposed to 𝑜𝑢𝑡𝑝𝑢𝑡

Efficiency= ∗ 100%
𝑖𝑛𝑝𝑢𝑡
operate reliably with the help of 52 weekly routine
preventive maintenances while system operated at Input ∝ maintenance cost.
100% efficiency. However, since the actual Thus, when maintenance cost is reduced the
operation was at 79% efficiency, then maintenances input cost will be reduced that is the overall
can be diligently optimized.100% utilization of the production cost reduced.
design production system would require 52  Increase efficiency of production line.
maintenances. Meaning 79% utilization of the 1
production system would require: 79%*52 = 42 Efficiency ∝ (4)
𝑖𝑛𝑝𝑢𝑡
maintenances. 10 saved maintenances time will reduce cost of
Therefore, maintenance thus reducing production cost
42 maintenances is the optimal number of (input) which in turn will increase the
maintenances which was required to keep the efficiency.
production system reliable. Some maintenances  Shift the production capability towards
thus, could be saved yearly from the actual design designing 100% efficiency. Though
number maintenance alone however much it can be
Extra maintenances = 52-42 =10 optimized cannot by itself shift efficiency to
= 10 maintenances (maintenances that can be 100% the shift would be towards that direction
saved). resulting from an optimized maintenance
system.
4.5. Saved Maintenance  Extend useful lifespan for CTC rollers. Machine
parts which get worn out during operation and
A usage-based maintenance is therefore, suitable needs to be changed during maintenance will be
to optimize further the existing periodic saved from being changed unnecessarily before
maintenance when necessary. The actual utilization fully utilised. CTC rollers are the main parts
of production line should be monitored by keeping which are changed during maintenance and
records of the tea handled daily and as the usage later re-sharpen for next use. So reduction of
approach to the 50000*6kg (300000kg) weight limit maintenance time would reduce wear and tear
before maintenance, then maintenance should be involved thus also less sharpening which in turn
scheduled. This scheduling should be based on the result in their lifespan being elongated before
usage and not based on period interval only. being fully changed.
By having a good planned usage-based  Optimized mentainance will reduce downtimes
preventive maintenance scheduling of maintenance hence it will improve the reliability of the
can be at the right time and thus, cost, time, labor production line
and purchases related to maintenance are minimized  The reduction on breakdowns and general
to an optimal level. improvements of the performance of the
machine will improve the safety of the factory
4.6. Results of saving 10 maintenance and reduce machine breakdown induced
accidents.
Usage based preventive maintenance, therefore,  Reduce unnecessary purchases.
can help to: During maintenance some small spare parts and
 Save up to 20% number of maintenances that is other resources needed are purchased for use
10 maintenances. When planned well usage but when maintenances reduce automatically
based maintenance can avail cumulative time of this unnecessary purchases reduce too.
up to ten maintenance which is a profit to the
company in aspects of money, time avail and 5. Conclusion
reduction of waste resources.
 Generally, avail time for production. 10 This research was focused on reducing the cost of
maintenances saved is equivalent to 60 days of production by employing usage based preventive
work avail at a long-term period of one year if maintenance. The quantitative analysis confirmed
there were enough tea production in the farm that preventive maintenance can help optimized
during all seasons. production further compared to periodic
 Reduce cost of maintenance. The number of maintenance. The results and finding showed that
maintenances cut from 52 to 42 automatically the 39615kg actual daily average tea process is
reduces the cost which would have instead been different from the 50000kg design capacity which
used for unnecessary maintaining activities. was supposed to work with help of 52 maintenances
 Reduce overall production cost. The input a year. In order to improve this low efficiency to a
resources which are utilized to 79% would be better level, the results shows that economic input
reduced by cutting short the much higher cost of can be optimized through minimization, since much
maintaining. is invested to a potential of 50000kg while actual

Electronic copy available at: https://ssrn.com/abstract=3949300

 Evans et al. SSRN-TESConf (2021) 016

output is 39615kg. Based on the focus of the study, Acknowledgements

the data analysis confirms that if the usage-based
preventive maintenance is used in place of periodic The authors acknowledged the financial support
preventive maintenance we can realize a save of up from the People’s Republic of China, ministry of
to 10 maintenances. The results also shows that Commerce, Hunan University
since maintenance cost is an element of total
production cost, thus, a reduction in it would result References
in a direct reduction in overall production cost.
Generally, from the finding and strategy of [1] O. A. Bamiro; D. Nzediegwu; K. A. Oladejo; A. Rahaman;
performing the analysis in this study, it gives a basis A. Adebayo (2011). Mastery of Technology for Junior School
to analyze other existing organization for Certificate Examination. Ibadan: Evans Brothers (Nigeria
optimization through minimization of resources and Publishers) Limited.
maximization of outputs. The strategy can be used [2] Marlow, D.R., 2010. Sustainability-based asset management
in the water sector. In: Definitions, Concepts and Scope of
to optimize both product and service output profit
Engineering Asset Management. Springer, London, pp. 261–
making organization. Also, from the results of this 275
study we can conclude that though the strategy to the [3] Uhlmann E et al. Maintenance, Repair and Overhaul in
findings is very good and the results being good for Through-Life Engineering Services. In: Redding L, Roy R,
making informed decision, still there is some editors. Through-life Engineering Services. Cham: Springer;
limitation. One of the limitation of the results is that [4] Niebel, B. (1994). Engineering maintenance management
the findings is a direct representation of the Chebut (industrial engineering) (2nd Ed.). CRC Press.
[5] Wilson, A. (2002). Asset maintenance management: A guide
tea factory and the solution will directly apply to this
to developing strategy and improving performance. Industrial
factory meaning you cannot generalize the results. Press, Inc.2015. p. 129-156.
The other limitation is that since the element of [6] Biehl, W.H., Inman, J.A., 2010. Energy optimization for water
production is occurring in natural setup it’s difficult systems. J. Am. Water Works Assoc. 102 (6), 50–55
to replicate the data even for next season. [7] Turkmenler, H., Aslan, M., 2017. An evaluation of operation
and maintenance costs of wastewater treatment plants: gaze
wastewater treatment plant sample. Desalin. Water Treat. 76,
6. Recommendations
382–388.
[8] Wendland, A., Ozoguz, Y., 2005. Operation costs of
In general from the results and findings of this wastewater treatment plants. In: EMWATER E-Learning
study it shows that existing production lines and Course. European Union, Germany.
systems can be operating normally without failure [9] Tsagarakis, K.P., Mara, D.D., Angelakis, A.N., 2003.
and major complication, however, they can be Application of cost criteria for selection of municipal
operating differently from expectation and can go wastewater treatment systems. Water Air Soil Pollution. 142
(1–4), 187–210
unnoticed for long. Therefore I recommend [10] Verbert, K., De Schutter, B., Babuska, ˇ R., 2017. Timely
generally that every product or service organization condition-based maintenance planning for multi-component
including my case study should occasionally systems. Reliable. Eng. Syst. Saf. 159, 310–321,
perform an analysis test of their company in their [11] Hukka, J.J., Katko, T.S., 2015. Resilient asset management
previous operation period so as to make wise and governance for deteriorating water services
decisions and adjustment in the following period of infrastructure. Procedia Econ. Finance. 21, 112–119,
[12] Alegre, H., do Ceu Almeida, M., Retrieved from 2009.
operation.
Strategic Asset Management of Water Supply and
To every profit and non-profit making Wastewater Infrastructures.
organizations, I recommend them to put their focus [13] Van der Lei, T., Herder, P., Wijnia, Y., 2014. Asset
on the maintenance activities and strategies for Management: The State of the Art in Europe from a Life
further optimization of production cost. Cycle. Springer.
I strongly recommend also that, a further research [14] Merriam-Webster. Archived from the original on 9 March
to be done on labor, storage and handling cost in line 2018. Retrieved 9 May 2018.
with, further optimization of production.

Electronic copy available at: https://ssrn.com/abstract=3949300