is defined, 
ents were 
ort system 
area, main 
the major 
e such as, 
| obstacle 
al hazards 
'egulations 
the system 
eir ratings 
e high-risk 
  
2 w/ risk 
SS 
1 support 
her tiered 
GIS based 
l| on n-tier 
Kr 
  
  
  
which has 
e 2). The 
a database 
siness tier 
mplements 
valuations 
nction of 
> user. 
| of H&S 
1 MS SQL 
server 2008. Furthermore, spatial data is stored as Geo-database 
file on ArcGIS Server 9.3.1. Spatial queries are executed by 
using Arc-Object API meanwhile queries that related to H&S, 
are performed via ADO.NET. Geo-processing, mapping, 
geometry and geo-data applications are realized via the arc 
object API at the business layer. Presentation interface is 
designed by using Asp.NET 3.5 on the Visual Studio 2008 
development platform with ArcGIS Web ADF tools and full 
AJAX solutions. 
2.2 Database Modelling 
H&S and spatial databases were used independently in solution 
architecture of system in order to provide integration with other 
construction management information system. 
2.2.1 H&S Data: It contains activity based risk assessment 
data. Risk assessment is an essential part of the planning stage 
of any H&S management system. It basically evaluates the risks 
involved in the execution of activities to provide the managers 
with information necessary to address intervention measures to 
comply with associated regulations. 
Various risk assessment methods, which can be classified as 
qualitative, quantitative and semi quantitative, may be used 
depending on the type of risk that is being considered and 
availability of data about the risk (Grassi et al., 2009; Huges 
and Ferret, 2007; Rowlinson and Lingard, 2005). In this 
research study, qualitative method was selected since it is a 
commonly used for risk assessment in pipeline projects. This 
method is appropriate where the level of risk does not 
correspond to the cost involved in applying a more detailed 
analysis. 
To perform a qualitative risk analysis, risk matrix method was 
used. In risk matrix, risks are rated according to the probability 
of their occurrence and their possible consequences (Table 1). 
Ratings can have a scale of three or five while the former is 
more commonly used. Probability and consequence are rated 
using verbal descriptors (e.g., medium-frequent probability and 
major severity) and cross referenced to establish the position of 
a risk in the matrix (e.g., 1 for low-seldom and 3 for major). 
These positions indicate the magnitude of the risk (e.g., 2 x 3 = 
6, high priority action), which can then be used to guide the 
selection of appropriate risk control methods and to establish 
priorities for the implementation of these controls. The greater 
the magnitude of the risk, the more effort should be put in its 
control, and the more urgently risk control actions should be 
implemented. 
  
  
Probability / Consequences / Risk Rate 
Likelihood Severity 
1 - Low 1- Slight (off work for 1 - No action 
(seldom) < 3 days) 
2 -Medium 2 - Serious (off work 2 - Low priority 
(frequently) — for» 3 days) action 
3 - High 3 - Major 3-4- medium priority 
(certain or (death/major harm) action 
near certain) 
6 - High priority 
action 
9 - Urgent action 
Table1. Probability, consequence and risk rate values used in 
the risk matrix 
An example of risk rating and mitigation measures is given for 
trenching activity in Table 2. 
  
  
Hazards Associated Risk Mitigation Measures 
Risks Rate 
Unsafe Injury to 4 - Ladders to be secured 
access, personnel at the top and 
egress and extended at least one 
falling into meter over the top of 
excavation the trench 
- Foreman checks the 
ladders and barriers at 
each shift. 
- Barrier excavations > 
4 m with appropriate 
fencing 
Table2. An example of risk rating performed for trenching 
activity based on the H&S standards 
In the proposed system, the hazards were extracted from H&S 
regulations, formalized and stored in database as interoperable 
with the GIS system. During the formalization of the hazard 
data, it was identified that there are two types of hazards: (1) 
Hazards which have risk ratings that are independent of 
projects, for example, a hazard that is related to breaking down 
of a stone crusher has a fixed risk rating, regardless of the 
project and location, (2) Hazards which have risk ratings that 
change based on project characteristics. Hazards that are 
dependent on the project characteristics are also grouped into 
two categories: (1) Spatial hazards, which can occur if some 
spatial characteristics exist, (2) non-spatial hazards, which are 
independent of spatial characteristics. 
2.2.2 Spatial Data: There are two fundamental approaches 
for representation of spatial data; vector model and raster 
model. In this paper study spatial analysis based H&S risk 
assessment study is addressed. The vector model allows us to 
represent specific spatial locations explicitly and provides the 
precise position of features in space. Based on analytical 
geometry, a vector model builds a complex representation using 
primitive objects such as points, lines and areas. The raster data 
model quantizes or divides space as a series of packets or units, 
each of which represents a limited, but defined amount of 
earth's surface. The raster model divides the earth into 
rectangular building blocks as grid cells or pixels that are filled 
with the measured attribute values. The location of each cell or 
pixel is defined by its row and column numbers. If the 
reasoning mechanism for identifying a spatial hazard is based 
on geographic objects represented by points, lines and polygons 
on the map (e.g., roads, underground cables), related hazard was 
represented in a vector model. The hazards that needs to be 
defined based on slope and altitude were represented in a raster 
model since they are associated with heights which is 
represented in raster data format in GIS.’ 
In the study geometric data was generated from 1:5.000 scaled 
topographical maps and 1:5.000 scaled pipeline layouts which 
are two fundamental resources for pipeline construction 
projects. These data resources are shown on Figure 3. and 
Figure 4. With respect to HS risks, relevant to terrain, are 
determined by spatial terrain analysis. Therefore it is based on 
raster data and its derived forms like height and slope data. 
Derived raster data and its source shown on Figure 6.