Total Cost of Ownership Analysis of 60 MVA

150/120 kV Power Transformer

Agus Indarto Rudy Setiabudy

Electric Power and Energy Studies, Department of Electric Power and Energy Studies, Department of
Electrical Engineering, Faculty of Engineering Universitas Electrical Engineering, Faculty of Engineering Universitas
Indonesia, Kampus Baru UI Depok 16424, Republic of Indonesia, Kampus Baru UI Depok 16424, Republic of
Indonesia & PT. CG Power Systems Indonesia, Kawasan Indonesia
Industri Menara Permai Kav. 10, Jl. Raya Narogong,
Cileungsi,Bogor 16820, Indonesia
Chairul Hudaya*
Electric Power and Energy Studies, Department of
Iwa Garniwa Electrical Engineering, Faculty of Engineering Universitas
Electric Power and Energy Studies, Department of Indonesia, Kampus Baru UI Depok 16424, Republic of
Electrical Engineering, Faculty of Engineering Universitas Indonesia
Indonesia, Kampus Baru UI Depok 16424, Republic of *Corresponding author: [email protected]
Indonesia

Abstract— Power transformer (PT) is one of the most PT will incredibly affect to the overall performance of electrical
important parts in electrical power system. In some cases, many power system. Thus, failure to maintain the PT will result in
electric utilities do not consider to capitalize the power losses, significant lost and degrade the reliability of electrical power
resulting in uneconomical decision during the purchasing. In this system [2].
study, to evaluate the best purchase choice of a 60 MVA 150/129
kV PT, we examine the total cost of ownership (TCO) as the basis In recent years, many electric utilities demanded the PT
for determining the total economic value of PT. When purchasing specifications to be more environmental friendly and more
and designing the PT, the optimum parameters are not only solely optimum in term of cost and technical aspects. For this reason,
determined by the design and the lowest cost, but also influenced calculating of total cost of ownership (TCO) of PT becomes an
by material cost, power losses and its typical operation. Hence, important factor to consider as this makes the buyer or user
design optimization and calculation of TCO of PT is of importance comprehensively understand the direct and indirect cost
to provide a reliable and economical product. Here it is realized associated with a certain product [3]. There are many users and
by reviewing the existing design and fabrication as well as the cost manufacturers consider only for the initial purchase price of PT
structure, design optimization and total cost ownership in their decision-making, rather than taking into account the
calculation. The result shows that the PT design optimization technical, economical and operational aspects. As a
employing losses capitalization exhibits the most economical value, consequence, the purchased PT was not as-desired one [4].
reducing TCO up to 3%. We found that the losses capitalization
and TCO of PT are sensitively affected by interest rate, economic This paper will address the following research-problem
life of PT, electricity cost, load factor and losses. Those factors are questions: (1) how to optimize the design of PT to achieve the
key components to determine the suitable specification, design most optimum cost? (2) how to calculate and achieve the most
evaluation and operation of the transformer. This practical study economic total cost of ownership in line with PT operation? and
is beneficial both for the owner and manufacturer of PT, thus (3) what consideration to be taken and what is the critical factor
giving the right specification and design to support purchasing in designing the PT? These critical questions are utmost
process of PT. important to optimize and evaluate TCO of PT as a base for
determining the optimum specification in PT purchasing
Keywords—total cost of ownership; power transformer; optimum
process.
design; losses capitalization;

I. INTRODUCTION
Power transformer (PT) is one of the most important parts in II. METHODOLOGY
electrical power system with the functions of stepping-up-and- We use a case study on PT design of 60 MVA 150/20 kV.
down the voltage, delivering electricity from the power plant to The evaluation was carried out by design optimization, losses
the consumer. Due to its critical function, PT acquisition capitalization and TCO calculation. We limited the scope of this
significantly contributes to the total electricity cost as it requires study to the existing data of PT design, material cost,
high initial capital cost [1]. On the other hand, the manufacturing transformer operational, cost of electricity and its financial
of PT takes considerable time owing to complicated procedures aspects. PT was designed and manufactured based on
to meet the high standard design. The reliability and quality of

978-602-50431-1-6/17/$31.00 ©2017 IEEE 291 2017 15th Intl. Conf. QiR: Intl. Symp. Elec. and Com. Eng
customized specification, applicable standard and (controlling the reactance), current density and cooling capacity.
manufacturer’s practice. Table 1 shows the specification of PT We further considered the no-load losses, full-load losses and
reference. noise level of the PT during the optimization. This study applied
two optimization approaches, namely optimization with fixed
TABLE I. SPECIFICATION OF PT REFERENCE (PT REF) no-load and load losses and optimization with evaluation of no-
load and load losses. The former approach will minimize the
Design Remark losses and cost, realized by varying the core diameter, flux
No-load losses (kW) 27 density, current density and cooling duct number. The goal of
Load losses (kW) 112 this approach was to meet guaranteed PT losses with a certain
Impedance (%) 12.57 value. Only losses as the pre-determined variable (input) was
Winding specs TV 226 mm2 (12.5 x 6.1mm) taken into calculation, while other parameters were optimized
Core specs 23SDR85, 0.8 W/kg
and analyzed. The later method minimizes the cost by varying
the core diameter, flux density, current density and cooling duct
Tank specs ST37
number. What makes different from the former scheme is that
Oil spec Uninhibited oil here the losses were capitalized.
IV. TOTAL COST OF OWNERSHIP (TCO)
III. OPTIMIZATION PROCESS
TCO calculation involves the costs of initial PT acquisition
We employed a case study for 60 MVA 150/20 kV PT and and operational costs to maintain the PT lifetime. The purpose
assessed its design optimization, losses capitalization, and of the TCO analysis is to minimize overall cost of PT, while
material selection to determine the optimum framework for PT increasing its possible lifetime, which is performed by
design. Variable input was given to the target design of PT minimizing the Equation 1.
reference. In consisted of general input such as delivery time
that affects to material price and technical input parameters ‫ ܥܱܶ݊݅ܯ‬ൌ ‫݊݅ܯ‬ሾ‫ ܶܥ‬൅ ‫ܲݔܣ‬଴ ൅ ‫ܲݔܤ‬௄ ሿ (1)
such as voltage and capacity, maximum oil temperature rise
(maximum guaranteed value at highest capacity), maximum where TCO is total cost of ownership (Rp), CT is the initial price
winding temperature rise (maximum guaranteed value at of transformer (tendered price, Rp), A is capitalization cost of
highest capacity), maximum hotspot temperature rise no-load losses (Rp/kW), P0 is no-load losses (kW), B is
(maximum guaranteed value at highest capacity), hotspot capitalization cost of load losses (Rp/kW) and PK is load losses
factor, the thickness of cooling duct (average), radiator type and (kW).
number, noise level, transformer tank thickness and type and Here CT was calculated as follow:
maximum transformer weight. Meanwhile, the economic
parameters for optimization include currency, core material ‫ ܶܥ‬ൌ 
஼்ெ
ൌ
஼ெெା஼ோெା஼ಽೌ್
(2)
price, winding (cooper) material price, transformer oil price, ଵିௌெ ଵିௌெ
insulation material price, transformer tank price and labor cost.
where CTM is PT manufacturing cost (Rp), SM is selling margin
(%), CMM is main material cost of PT (Rp) such as core,
winding, oil, tank, insulation, and cooling system, CRM is other
material cost (Rp) such as protection and control system, on load
tap changer and CLab is production or labor cost (Rp).
The steps to calculate the TCO of PT has been outlined in [6]
as follows: (1) calculating the material cost of transformer (CT
component); (2) calculating the manufacturing cost (CT
component); (3) calculating the selling price of transformer
(CT); and (4) calculating the TCO of PT. The capitalization
losses cost is analyzed by net present value (NPV) of electricity
cost [7-8] as follow:

஻௉௉௫ሺଵା௞ሻ೙షభ
‫ ܣ‬ൌ ͺ͹͸Ͳ‫ ݔ‬σ௡௡ୀଵ ቀ ቁ (3)
ሺଵା௜ሻ೙

஻௉௉௫ሺଵା௞ሻ೙షభ
‫ ܤ‬ൌ ͺ͹͸Ͳ‫ ݔ‬σ௡௡ୀଵ ቀ ቁ ‫ݔ‬ሺ಺಺ಳሻଶ  (4)
ሺଵା௜ሻ೙ ೃ
Fig. 1. Design Optimization Process of Power Transfomer
஻௉௉௫ሺଵା௞ሻ೙షభ
The optimization procedure, starting with existing PT design ‫ ܥ‬ൌ ͺ͹͸Ͳ‫ ݔ‬σ௡௡ୀଵ ቀ ቁ ‫ݔܨܨݔ‬ሺ಺಺ಳሻଶ (5)
ሺଵା௜ሻ೙ ೃ
as reference is shown in Fig.1. The PT reference was with fixed
winding configuration and comparable both voltage and
where A is no-load losses cost, B is load losses cost, C is cooling
impedance. The optimization process includes the main
fan losses cost, n is transformer economical life (25 years), BPP
parameters such as core diameter, flux density, winding height
is cost of electricity, k is incremental factor of BPP, i is interest

292
rate (%), IB is load current (A), IR = nominal current (A), and FF
is PT capacity (p.u) where the cooling fans start to operate (0.6).
The optimization results for each scenario is shown in Table
As described in Equation 3 and 4, component A and B are 3.
unique factors for PT owners that are depending on PT loading
and cost of electricity. PO and PK are technical parameters of PT TABLE III. COMPARISON OF OPTIMIZATION DESIGN AMONG THE SCENARIOS
representing no-load and load losses respectively.
PT
Parameter D1 D2 D3 D4
TCO calculation was then performed with a reference to Ref.
Equation 3, 4 and 5 for transformer lifecycle. This calculation No-load losses (kW) 27 30(*) 25(*) 20 31
Load losses (kW) 112 115(*) 110(*) 253 124
was based on cost of electricity and transformer loading, as well Total losses (kW) 139 145 135 273 155
as an estimated energy price and load growth. Interest rate and Impedance (%) 12.57 12.5 12.5 12.5 13
economic life was also included to precisely determine TCO. In Core weight (%)(1) 100 97 103 63 96
general, TCO components is displayed in Fig.2. Winding weight (%)(1) 100 85 113 44 72
Oil weight (%)(1) 100 94 105 67 89
Tank weight (%)(1) 100 95 104 67 91
Radiator weight (%)(1) 100 108 95 256 116
Total weight (%)(1) 100 93 106 64 88
Material cost (%) (1)(2)(3) 100 91 119 57 88
Cost of losses (k $)(2)(4) 580 609 560 1,069 648
Cost of losses (%)(1)(2) 100 105 97 183 112
Total Cost (%) (1)(2)(5) 100 99 102 130 97
Notes :
(*) Losses value as input for optimization, (1) Value in the table is comparison with design reference.(2)
Material cost describe initial purchase cost of transformer, meanwhile total cost describe initial cost and
losses cost during transformer operation. (3) Material cost is material and labor cost which depend on
weight of material and unit material cost. (4) Losses cost was calculated as follow: losses cost = A x PO
+ B x PK with Factor A = 5,874 $/kW and Factor B = 3,759 $/kW and (5) Calculation results ($), total
cost is the sum of material cost (1) and losses cost (2).

Fig. 2. Components of Transformer TCO The highest material cost of PT was achieved by the lowest
load losses and vice versa. We found the optimum design was
V. RESULTS AND DISCUSSION exhibited by scenario D4 (97%), postulating the lowest losses
The PT optimization was conducted by taking the PT capitalization compared to PT reference and other scenarios.
reference of 60 MVA 150/20kV, following four (4) scenarios: When the PT purchase only consider initial cost of PT
acquisition, the total cost can be decreased up to 57% of PT
1) Scenario D1: optimization with fixed losses in which the reference cost. The reason is because the optimization
losses value of observed PT is higher than that of PT calculation consider only for material cost. In contrast, the
reference losses (PO and PK of observed PT > PO and optimization with losses capitalization resulted in a decreased
PK of PT reference) TCO for about 3% from PT reference.
2) Scenario D2: optimization with fixed losses in which the For TCO calculation we employed 5 scenarios as shown in
losses value of observed PT is lower than that of PT Table 4. The calculation process follows the procedures as
reference losses (PO and PK of observed PT < PO and outlined in Fig.3.
PK of PT reference).
TABLE IV. SCENARIO TCO CALCULATION
3) Scenario D3: optimization with losses evaluation and
no-losses capitalization. Scenario Interest Cost of Loading Economic
Rate (%) Electricity Factor (%) Life (years)
4) Scenario D4: optimization with losses evaluation and (Rp)
losses capitalization (including both no-load and load PT Ref. 10 890 and 1% 80 25
losses cost (scenario D4) increase
anually
In this study we assume some parameters based on actual TCO 1 Interest rate is changed, other parameters are fixed
price applied by PT manufacturer in Indonesia. Table 2 shows TCO 2 Economic life is changed, other parameters are fixed
TCO 3 Cost of electricity is changed, other parameters are fixed
our assumption for this study. TCO 4 Loading factor is changed, other parameters are fixed
TCO 5 Load losses is changed, other parameters are fixed

TABLE II. UNIT PRICE/JOB COST ASSUMPTION

Materials Unit Price ($/kg)
Core 3.13
Winding (copper) 8.23
Transformer oil 1.03
Insulation 3.98
Transformer tank 3.13
Labour cost 2.85

293
 cost of electricity will increase the TCO for about 9%, meanshile
a decreased 20% of that will decrease TCO for about 9%. A 20%
increased economic life will increase 2% TCO, in opposite a
20% decreased will decrease 3% TCO, as shown in TCO3. For
TCO4, an increase load factor of 20% will increase TCO of
19%, a decrease of that value will decrease TCO for about 15%.
This is due to the load losses during operation. Finally for TCO5,
a 20% increase and decrease of load losses will increase and
decrease TCO for about 4%, respectively.
Sensitivity analysis was then carried out to assess the effect
of input change to TCO values as shown in Fig. 4.

Fig. 3. The procedures for TCO calculation

TCO calculation results for each scenario was shown in

Table 5.
Fig. 4. Sensitivity Analysis of TCO
TABLE V. CALCULATED TCO
From sensitivity graph we can conclude that interest rate and
A B C CT CC TCO
PO PK BPP N I FB
(million (million (million (billion (billion (billion load losses have negative correlation with TCO. Meanwhile,
(kW) kW) Rp/kW (yr) (%/yr) (%)
Rp. /kW) Rp. /kW) Rp. /kW) Rp.) Rp.) Rp.) economic life, cost of electricity and load factor provide positive
TCO 1 impacts to TCO for about 20%.
31 124 890 25 8 80 90.5 57.9 34.8 9.6 10.2 19.8
VI. CONCLUSION
31 124 890 25 10 80 76.4 48.9 29.3 9.6 8.6 18.2
31 124 890 25 12 80 65.5 41.9 25.2 9.6 7.4 17.0
In this study, the TCO of 60 MVA 150/20kV PT was
TCO 2 investigated by taking the technical, economical and operational
31 124 890 20 10 80 70.9 45.4 27.2 9.6 8.0 17.6 aspects into account. We found that the PT optimization without
31 124 890 25 10 80 76.4 48.9 29.3 9.6 8.6 18.2 considering the power and capitalization losses contributed to
31 124 890 30 10 80 79.9 51.2 30.7 9.6 9.0 18.6 the total losses of 253 kW and exhibited to the lowest material
cost of 57% of PT reference. Nevertheless, it provided the
TCO 3
highest total losses cost of 183% of PT reference. With this
31 124 712 25 10 80 61.1 39.1 23.5 9.6 6.9 16.5 scheme, total cost (including material + total losses cost)
31 124 890 25 10 80 76.4 48.9 29.3 9.6 8.6 18.2 achieved 130% of PT reference. In other scenario, the PT
31 124 1068 25 10 80 91.6 58.7 35.2 9.6 10.3 19.9 optimization without determining the power losses, but
TCO 4 concerning the load losses, gave the total losses of 155 kW and
benefited to the lowest total cost of 97% of PT reference. These
31 124 890 25 10 60 76.4 27.5 16.5 9.6 5.9 15.5
facts suggested that the PT optimization is about the trade-off
31 124 890 25 10 80 76.4 48.9 29.3 9.6 8.6 18.2 between PT materials and losses cost, in order to realize the
31 124 890 25 10 00 76.4 76.4 45.8 9.6 12.1 21.7 optimum total cost. We also observed that the lowest TCO was
TCO 5 accomplished at highest interest rate, lowest electricity cost, and
31 100 890 25 10 80 76.4 48.9 29.3 11.5 7.4 18.9
lowest load factor. Other parameters positively correlated with
TCO of PT are electricity cost, economic cycle-life and load
42 124 890 25 10 80 76.4 48.9 29.3 9.6 8.6 18.2
factor. In contrast, both interest rate and load losses negatively
31 148 890 25 10 80 76.4 48.9 29.3 7.7 9.7 17.4 affect the TCO of PT.
ACKNOWLEDGEMENT
For TCO1, the incease of interest rate 20% will decrease
The authors would like to thank the financial support
TCO for about 7%. In contrast, a decrease 20% of interest rate
provided by Universitas Indonesia through the 2017 PITTA
will increase TCO for about 9%. For TCO2, an increase 20% of

294
funding scheme managed by the Directorate for Research and
Community Engagement (DRPM) Universitas Indonesia.

REFERENCE

[1] Considerations for A Power Transformer Emergency Spare Strategy for

The Electric Utility Industry, The Electric Power Research Institute for
the U.S. Department of Homeland Security Science and Technology
Directorate, September 30, 2014
[2] James H. Harlow, Power transformer engineering / editor, -- 2nd ed. p.
cm, ISBN-13: 978-0-8493-9186-6 (alk. paper)
[3] Considerations In Application And Selection Of Unit Substation
Transformers, Charles J. Nochumson, IEEE IAS Pupl and Paper Industry,
Portland 2001
[4] A Return on Investment as a Metric for Evaluating Information Systems:
Taxonomy and Application, Alexei Botchkarev, Peter Andru,
Interdisciplinary Journal of Information, Knowledge, and Management
Volume 6, 2011
[5] PT CG Power System Indonesia, Design Manual, Design & Technology
Department, 2016
[6] Pavlos S. Georgilakis, Spotlight on Modern Transformer Design, Springer
Dordrecht Heidelberg 2009, ISBN 978-1-84882-666-3.DOI 10.1007/978-
1-84882-667-0
[7] Draft SPLN T3.007-2: 2016, Spesifikasi Transformator Tenaga Bagian 2:
Transformator 150/22/10 kV, Draft Agustus 2016
[8] PT PLN (PERSERO), Buku Standardisasi Spesifikasi Teknis
Transformator Tenaga, 0632.K/DIR/2013, 17 Oktober 2013
[9] R. P. W. S. Amarasinghe, W. G. K. P. Kumara, R. A. K. G. Rajapaksha,
R. A. D. K. Rupasinghe and W. D. A. S. Wijayapala, "A transformer
design optimisation tool for oil immersed distribution transformers,"
Moratuwa Engineering Research Conference (MERCon), 2015,
Moratuwa, 2015, pp. 100-105.

295