 Location-Allocation Problem for Blood Service Facility in Thailand
Pornpimol Chaiwuttisak1*, Dr Honora Smith1, Dr Yue Wu1, Tassanee Sakuldamrongpanich2, Dr Somchai Pathomsiri3
1 CORMSIS,
University of Southampton,
2
National Blood Centre, the Thai Red Cross Society, 3 Civil and Environmental Engineering department, Mahidol University
* Corresponding author. e-mail address (): pc12g09@soton.ac.uk
Problem Background
Introduction
In principle, blood service centres have been established to
 SAME QUALITY STANDARD AND ASSURE BLOOD SAFETY 
The Thai Red Cross Society (TRCS) has difficulty
building new blood centres (see in Figure 1) with full
functions because setup costs are expensive. Current
amounts of donated blood are insufficient to meet
demand. More fixed collection sites are proposed to
make easy access for donors. Some sites may perform
preparation and storage for blood that hospitals can
receive directly. However, location decisions have
impact on the blood supply chain and logistics.
Furthermore, the investment budget is limited for this
non-profit organisation.
There is one National Blood Centre (NBC) at Bangkok
and twelve Regional Blood Centres (RBCs) in the
different provinces over the country.
Figure 1
Objectives of Study
We present an Integer Programming model for a location-allocation
problem, “Max donor supply – Min distribution distance”, based
on two classical facility location models:
Maximum Covering model + p-median model
The weighted objectives of the model minimise the demand weighted
distance to supply blood to hospitals and maximise the quantity of
expected blood donors from volunteers in order to find the optimal
sites for two kinds of facilities (see in Figure 1) to serve blood:
Formulation
Results and Discussions
Figure 2
Minimise
𝑤1
𝐽
𝑤2
𝑖
1
𝑘 𝑧𝑗𝑘 𝑑𝑗𝑘
1
𝑘 𝑥𝑖𝑘 𝑑𝑖𝑘
+
𝐽
𝑢𝑖 +
1
𝑦
𝑗 𝑗 𝑠𝑗 +
−1 × 𝑤3 (
2
𝑘 𝑧𝑗𝑘 𝑑𝑗𝑘 +
2
𝑖 𝑗 𝑥𝑖𝑗 𝑑𝑖𝑗 𝑢𝑖
2
𝑦
𝑗 𝑗 𝑠𝑗 )
Figure 3
+
Subject to
𝑘
𝑘
𝑘
1
𝑥𝑖𝑘
+
1
𝑧𝑗𝑘
2
𝑥𝑖𝑗
=1
𝑗
1
𝑦𝑗
∀𝑗
2
𝑧𝑗𝑘
= 𝑦𝑗2
∀𝑗
=
2
𝑥𝑖𝑗
≤ 𝑦𝑗2
1
1
𝑧𝑗𝑘 ≤ 𝑦𝑗
2
2
𝑧𝑗𝑘 ≤ 𝑦𝑗
𝑗
∀𝑖
∀𝑖 ∀𝑗
∀𝑗 ∀𝑘
∀𝑗 ∀𝑘
𝑡𝑐𝑗1 𝑦𝑗1 +
𝑦𝑗1 + 𝑦𝑗2 = 0,
𝟏
𝒙𝒊𝒌 ,
𝑗
𝑡𝑐𝑗2 𝑦𝑗2 ≤ 𝑏𝑢𝑑𝑔𝑒𝑡
𝑗 ∈ 𝑠𝑒𝑡 𝑜𝑓 𝑏𝑙𝑜𝑜𝑑 𝑐𝑒𝑛𝑡𝑟𝑒𝑠
𝟐
𝒙𝒊𝒋 ,
𝟏
𝒚𝒋 ,
𝟐
𝒚𝒋 ,
𝟏
𝒛𝒋𝒌 ,
𝟐
𝒛𝒋𝒌
where
denote
decision variables ∈ [0,1] for type 1
(collection site) and type 2 (collection and
distribution site), 𝒕𝒄𝟏𝒋 and 𝒕𝒄𝟐𝒋 are total of
fixed costs, 𝒖𝒊 denotes blood usage of
hospital 𝒊, 𝒔𝒋 denotes quantity of donated
blood from the site 𝒋, 𝒘𝟏 and 𝒘𝟐 and 𝒘𝟑
are weight scores for objectives.
Acknowledgement 
We acknowledge the support of

We
use
equally
weighted
objective
components after discussion with the National
Blood Centre. For lower budgets, up to 60
million Baht, only the less costly rooms for
collection only are located. At higher budget,
the more costly rooms for collection and
distribution also can be located. In particular,
one site in the northern region, Chiang Rai,
and one site in the southern region,
Chumphon, are recommended as distribution
centres illustrated in Figure 3. From Figure 2,
values of the first objective decrease while the
others increase when budgets are higher.
Future work
Future work will be to study in detail the
logistics of blood supply in one particular
region, in the north of Thailand.
References 
William P. Pierskalla (2004), Supply chain management of blood banks, in: M.L. Brandeau, F. Sainfort, W.P. Pierskalla(eds.), Operations Research and
Health Care: A Handbook of Methods and Applications, Kluwer Academic Publishers, 2004, 103-145
1) The Royal Thai Government in providing a Ph.D. scholarship for this research
2) National Blood Centre, The Thai Red Cross Society in providing data