Transportation Planning and Simulation of Petroleum Products to Area Struck by Tohoku Japan Earthquake

■2012 JSPS Asian CORE Program, Nagoya University and VNU University of Economics and Business

Transportation Planning and Simulation of Petroleum Products to
Area Struck by Tohoku Japan Earthquake

Nagoya University Norie Inagaki*
Nagoya University Yan Liu*
Nagoya University Soemon Takakuwa*

ABSTRACT: Tohoku Japan Earthquake, happened in March 11, 2011, causes great destruction to our supply system of
crude oil and petroleum products, and produces great influence. As main road, refinery, oil-tank terminal and gas station
are all destructed and no effective solutions are followed as soon as earthquake happened, supply system can not be
guaranteed immediately. Some summary is made including our supply chain of petroleum products, supply and support
system of petroleum products after earthquake, and transportation of petroleum products to refuges. Also analysis
validation of transportation of gasoline and kerosene to evacuation sites are made.
KEYWORDS: Tohoku Japan Earthquake, Transportation planning, Oil products

1. Foreword
The disaster due to Tohoku Japan Earthquake and
tsunami are far more serious than government and
Petroleum Association of Japan expect, especially to
our petroleum supply system. We are suffering great
and disastrous results. Locations and equipments such
as petroleum refining factory, tank, gas station, and SS
(service station) are all destroyed heavily. Meanwhile,
as no effective solutions are made when roads are
blocked, transportation are cut off, our supply system
could not be guaranteed immediately. So, through
review and summarization to our daily supply chain of
petroleum products, supply and support system right
After disaster, and transportation of petroleum products

to refuges will be made, and then a introduction of
transportation system for petroleum related products to
evacuation site in Tohoku region. A linear
programming model is also constituted especially for
transportation of petroleum and refueling to evacuation
site. With solid analysis and verification, a study of
solutions for future disaster could be conducted.

by tank. Petroleum refining and future processing are
conducted in 27 refining factories nationwide. Refined
petroleum are well stored in about 190 oil terminals
and transported all over the country for sale with
coasting tanker, tank car and tank truck when necessary.
Fig. 1 shows the detailed information about petroleum
supply chain.

Fig. 1 Supply chain of oil and petroleum products
(Source: PAJ)
In one year, about 17 million KL crude oils are
imported from foreign countries used by tankers.
About 78 thousand KL crude oils are produced in
Japan. Also, about 18 million KL crude oils are saved
or refined.

2. Supply chain of petroleum products
Petroleum products are imported by tanker, and stored
*Nagoya University, Graduate School of Economics and Business Administration


3. Petroleum products supply and support system

after disaster
Based on press release by Ministry of Economy, Trade
and Industry, the article makes the summary of
petroleum products supply and support system. It is
since March 17 that Ministry of Economy, Trade and
Industry began to public specific solutions, and issued
focus on March 21.

pumped through oil drum to cars or can one carried.
Another advantage of oil drum is no need of tank truck
but transported with ordinary truck. So in Miyagi Pref.,
Iwate Pref. and Fukushima Pref. where SS operation
rate is relative low, oil drum are provided in 10 to 20
refuges.

3.1 Restoration of SS (service station)

On March 27, specially erected mini SS No. 1 is set in
5 places in Rikuzen Takata City for refueling. On April
1, it is expanded to 9 cities towns and villages [1].

Service Station (SS is called for short) is a facility that
supply petroleum products for people. SS has a
function that people can obtain petroleum products
conveniently.

As of May 31, specially erected SSs are closed. There
are 10300 specially erected SSs making refueling
supply. For detailed number of specially erected SSs,
please refer to Fig. 3.

Primary oil distributor‟s service station is 3070 in
Tohoku region. The news is first published on March
23 that No. of SS still available is 2265, and operation
rate is 74%. Till March 30, every petroleum enterprises
release its SS situation through official internet. On
April 4, SS‟s operation rate is increased to 90%, and
then no marked rise until May 30 that operation rate is
up to 96%. However, influenced by aftershock, a
drastic decline to 78% happens on April 8. SSs remain
unstable. Fig. 2 shows process of SS restoration.
Fig.3 Number of specially erected service stations
(source: METI)
3.3 Destruction to refining stations and actions of
nation
Fig. 4 shows capability of crude oil refining and
recovery of refinery. After the disaster, 31% of
refineries in Tohoku region and Kanto region along
Pacific Ocean lose productivity. It is until March 21
that 17% of refineries begin to produce.
Fig.2 Restoration of service stations
(Source: METI)
In Kanto region, 88% of 5838 are available within
6661 series on March 28, and 99% of 6618 recovered
for use on April 11.

The Refining Ability of Oil Wholesaler and
The Restored Situation of damaged Oil Factories

out of operation

in operation

69％

１４％

Damaged Oil
Factoris

restart by 21st,
March

17％

3.2 Specially erected SS (set in refuges)
In primary and mid school refuges, oil drums are used
for petroleum and refueling container. Emergent supply
guarantees that victims can be provided with small
amount of storage. By cooperation with petroleum
enterprises, timely and continuous supply can be
guaranteed within a period. Capacity of each oil drum
is approximately 200L. One refuge with about 100 oil
drums and 20KL capacity of oil in maximum. Oil is

Refining Ability of Whole Japan
(as of January, 2011)
718,000 (kiloliter per day)

Fig.4 Restoration of refineries of primary oil
distributers (Source: The Nikkei)

On March 14, Japanese government decided to reduce
private oil stock from 70 days amount pre-disaster to
67 days amount post-disaster. With the implementation
of focused solutions, oil stock is further reduced to 45


days amount. The work ended by May 20 and the
amount of oil stock resumed to 70 days.
On March 17, Ministry of Economy, Trade and
Industry public their solutions of private oil stock
change.
1. Tohoku region
(1) Additional supply of tank truck from refineries in
western Japan to Tohoku region.
(2) Function recovery of oil terminals along Pacific
Ocean; Through SS and emphasized supply
(3) Special solutions for regions out of supply.
(4) Guarantee the railway transportation; answer the
request from evacuation sites; support to surroundings
of Fukushima nuclear power station.
(5) Build special erected SS in evacuation sites; Kanto
region
(6) Transportation from western Japan; stock crash in
available
2. Kanto refineries
(1) Corporation between entrepreneurs and cunning
supply system; appoint emergent SS and emphasized
supply.
(2) Normal railway transportation to Utsunomiya,
Takasaki, etc.

4. Transportation planning for petroleum products
to evacuation sites
4.1 Formulation of transportation planning
By measures mentioned in chapter 3 which carried out
Ministry of Economy and Japanese government, the
situation of evacuation sites in Tohoku Japan was
improved little by little. However, after the Tohoku
Japan Earthquake, because of lack of refueling, no oil
stove could be used and victims are suffering torture of
severe coldness. Also for insufficient supply of
petroleum for evacuation site, car use is limited. We
can often see reports that through search for families
could not be conducted because of car limitation. After
the earthquake, all refineries, oil terminals and SS in
Tohoku Japan are damaged. It becomes difficult to
transport petroleum products to evacuation sites. There
are urgent needs to transport petroleum and refueling
from non evacuation sites to refuges. It is our most
urgent task to work out plans for amounts and means
of petroleum and refueling transported from non
evacuation site stock to refuges. In urgent and special
period of time, it is better to use truck transportation
with oil drum of 200L capacity.
Above problems could be solved by formulation with
Linear Programming. LP is a mathematical approach
with which limited resources can be allocated
effectively [3]. After the Tohoku Japan Earthquake on

Mar. 11, LP is used in evacuation sites to make
emergent petroleum and refueling transportation plan

and flexible petroleum products supply plan.

T1

Oil
Terminals

2

1

T2
T2

4
3
9
T3

7
8
5

6
T4
10

11

13

12

14

Fig. 5: Map of oil terminals and the special service
stations
Fig. 5 shows locations of each oil terminals and special
SS (hereinafter “SS”), and possible roads available
after the disaster. Unavailable roads are showed with
orange lines. Available roads from each oil terminal are
showed with dark blue, green, purple and yellow lines.
Along Tohoku Japan coastline, there are only 4 oil
terminals free of damage. Petroleum and refueling are
drummed and transported with trucks to 14 SSs. Fig. 6
shows the necessary transportation time from oil
terminals to SSs. The four oil terminals are respectively
represented by T1, T2, T3 and T4. 14 SSs are
represented by S1 ~ S14. Numerical values marked on
nodes show necessary transportation time from oil
terminals to SSs. For example, it takes 2.5 hours from
T1 to S1. Demand of S1 is 0.63 kiloliter. Number
beside the oil terminal represents possible supply. For
instance, the possible supply of T1 is 80 kiloliter.
According to rules about the time from oil terminals to
SSs, demands of SSs and possible supplies from oil
terminals, we can make quantitative transportation
from oil terminals to SSs to satisfy their demands with
minimum transportation time.

Table1: Expected demand quantities of gasoline and
kerosene (kl/day)

Fig. 6: The time required for transportation from oil
terminals to SS stations
The formulation is constructed as follows.
Objective function:
4

i 1 j 1

( i = 1,2,…,4; j = 1,2,…,14)
Subject to:

x
j 1

ij

4

x
i 1

ij

 Ti

 Sj

(i = 1,2, …,4)
(j = 1,2,…,14)

xij  0 (i=1,2,…,4; j=1,2,…,14)

Evacuee number (peple)Gasoline demand（kl/day）Kerosene（kl/day）
709
0.28
0.35
10174
4.07
5.09
40455
16.18
20.23
1171
0.47
0.59
1836
0.73
0.92
4808
1.92
2.40
2409
0.96
1.20
15098
6.04

7.55
17738
7.10
8.87
3172
1.27
1.59
3115
1.25
1.56
1269
0.51
0.63
1339
0.54
0.67
2587
1.03
1.29

Table2: transportation time from oil tanks terminals to
SS stations (hours)

14

Z＝min  H ij xij

14

S1

S2
S3
S4
S5
S6
S7
S8
S9
S10
S11
S12
S13
S14

(1)
(2)
(3)

S1

S2

S3

S4

S5

S6

S7

S8

S9

S10

S11

S12

S13

S14

T1

2.50

3.50

5.00

3.00

5.00

5.50

4.00

5.50

5.00

5.50

6.50

7.00

9.00

7.00

T2

2.70

4.50

4.00

2.00

3.50

4.00

3.20

4.00

4.00

4.50

5.00

5.50

7.50

6.00

T3

4.00

6.00

5.50

3.00

2.50

2.50

3.00

4.20

5.00

3.00

4.00

4.00

6.50

4.50

T4

6.00

8.50

7.00

5.50

4.00

3.50

4.50

5.00

6.50

3.20

2.50

3.00

4.00

3.50

According to the LP function and data above, by using
computer technologies, the optimal solution is obtained
as shown in Table 3. So the optimal solution of Z (the
total sum of „transportation time‟ times „shipping
quantity‟) is 353.025.

Where,
H ij ：the necessary transportation time from oil tanks i
to SS station j (hours)
i: index number of oil tanks, and it is integer
j: index number of SSstation, and it is integer
S j : expected demand quantity to SS station j (kl/day)

Table3: Optimal solution of the decision variables

(kl/day)
S1

S2

S3

S4

S5

S6

S7

S8

S9

S10

S11

S12

S13

S14

T1

0.63

9.16

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

T2

0.00

0.00 36.41

1.06

0.00

0.00

0.00 13.59 15.97

0.00

0.00

0.00

0.00

0.00

T3

0.00

0.00

0.00

0.00

1.65

4.32

2.16

0.00

0.00

2.86

0.00

0.00

0.00

0.00

T4

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

2.81

1.14

1.21

2.32

Ti : available supply quantity from oil tanks i(kl/day)
xij ：shipping quantity from oil tank i to SS station j

(kl/day)
The object is to minimize the total sum of
„transportation time‟ times „shipping quantity‟. The
constraints are: (1) supply amount for SSs should not

be more than available amount oil terminals could
afford; (2) as a decision variable, supply from oil
terminal i to SS j must be non-negative.
4.2 Case study
Table 1 shows assumed demand of petroleum and
refueling from each SS. The assumption is based on
the facts that 5 persons take one car and consumes 2L
petroleum per day. Besides, we can also work out
demands of petroleum and refueling of SS based on
assumption that 10 persons use one camp stove.
From Table 2, we can know road situation after Tohoku
earthquake, and transportation time from oil terminals
to SSs.

It shows necessary amount of petroleum and refueling
transported from each oil terminals to mini SSs in
refuges. For example, the optimal amount from oil
terminal T1 to service station S1 is 0.63KL per day.
5. Simulation Analysis
5.1 Simulation Logic
In chapter 4, optimal amount of petroleum products
from oil terminals to SSs are solved. However, it is not
solved about the optimal number of the truck used for
transportation. This transportation problem for
earthquake and recovery is very important. So, in this
paper, a simulation model is applied to analyze this
issue.
Based on Table 2 and Table 3, results of LP solution is
embedded into the simulation model for inputs. And, a
simulation logic flow is constructed as shown in Fig.7.

Create Entity Trucks (loading
petroleum products）

Set the number of the trucks

n=1

Assign transportation time and
quantity of petroleum products

Operation

Read data (transportation
time and quantity)

Results of
LP

Simulation running
Check the data
(Do each data
accord?)

Results (the time for transport)

n=n+1

Input

Check
(Is the number of Entity Trucks
optimal?)
No

Optimal solution

Yes
End

Fig.7 Flow chart of the simulation model
In the Fig.7, it can be seen that the parameters are
inputted through ReadWrite model. By setting and
regulating the number of trucks, the total finished time
for the entire transportation system is obtained. By
changing the input information, this model can deal
with other transportation planning. In Table 3, the
longest time of LP optimal solution is 4 hours. When
the total time outputted from simulation is equal to 4
hours, the simulation running cycle is ended.
5.2 Simulation Result
Fig.8 shows the results of the simulation running. It is
proved that one, two or three trucks cost a lot of time
for transportation. With increasing the number of
trucks, the total finished time for the entire
transportation system is changed. But over fours trucks,
the time is not changed. It means that four trucks are
enough to the transportation system. In addition, the
total time for four or over four trucks is 4 hours that is

equal to the LP optimal solution. And each truck is
distributed into each oil terminals. It is not only proved
that the simulation is effective, but also that the optimal
number of trucks is achieved, which can save the
limited resource for earthquake and recovery area.
Fig.9 shows the simulation animation for optimal four
trucks.

Fig. 8 Result of simulation
6. Conclusion
The article mainly discusses petroleum products supply
after Tohoku Japan Earthquake, especially provides
linear planning programming to solve problem of
necessary transportation for evacuation sites. LP
provides a very effective approach to make most
favorable transportation plan within shortest time under
limited resources and time. It is expected that the
research results can be applied if next disaster happens
in future.


Note
[1] Iwate Prefecture: Kamaishi City, Rikuzen-Takata City,
Ohtsuchi Machi
Miyagi Prefecture: Ishinomaki City, Higashi-Matsushima
City, Onagawa Machi, Minami-Sanriku Machi
Fukushima Prefecture: Shinchi Machi, Iidate Mura

ACKNOWLEDGEMENTS
This research was supported by the Grant-in-Aid for Asian

CORE Program "Manufacturing and Environmental
Management in East Asia" of Japan Society for the
Promotion of Science (JSPS).

REFERENCE
[1]
[2]
[3]

[4]

Fig.9 Simulation animation for optimal four trucks.

/> />Ignizio, J. P., and T. M. Cavalier. (1994) Linear
programming. Englewood Cliffs, New Jersey:
Prentice-Hill.
R. P. Sadowski and D. T. Sturrock. (2003) Simulation
with ARENA Third Edition, Singapore: The
McGraw-Hill Companies, Inc.