Safety Science 50 (2012) 1941–1951

Contents lists available at SciVerse ScienceDirect

Safety Science

journal homepage: www.elsevier .com/locate /ssc i

Managing safety in hydropower projects in emerging markets – Experiencesin developing from a reactive to a proactive approach

Urban Kjellén ⇑NTNU, N-7491 Trondheim and SN Power, PO Box 200, Lilleaker, N-0216 Oslo, Norway

a r t i c l e i n f o a b s t r a c t

Article history:Available online 3 September 2011

Keywords:Hydropower construction projectsIndiaPhilippinesOwner interventionSafety performance

0925-7535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.ssci.2011.07.018

⇑ Tel.: +47 24068633; fax: +47 24068621.E-mail address: urban.kjellen@snpower.com

International companies investing in emerging markets need to address the management of safety undervarying conditions. The paper presents an evaluation of the approaches taken by an international hydro-power company in two different construction projects, one in India and one in the Philippines. In the firstproject, the company had to intervene as a reaction to poor safety and overall performance. Based on pre-vious experiences, the company selected a proactive approach in the second project, involving use of thecontracting process to ensure adequate conditions from the start of construction and onwards. The paperaccounts for the safety performance results in the two projects and interprets them in relationship toexternal and internal influencing factors in the respective project. It is concluded that a proactiveapproach has the potential of delivering satisfactory safety results at moderate costs for follow-up. Areactive approach risks exposing the company to ethical and reputation risks due to a poor safety perfor-mance before the corrective actions have taken full effect. It is also resource demanding. Results demon-strate the feasibility of turning around a project’s safety performance also when preconditions arechallenging.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Electric power is vital for increased wealth in developingcountries. In the two countries of current interest in this paper,India and the Philippines, there is a well-documented positivecorrelation between growth in electricity demand and in grossdomestic product (IHS Global Insight, 2010; CEA, 1999–2009).During 1999–2009, the gross domestic product in India rose by86% at the same time as the growth in electricity demand was61%. The corresponding figures for the Philippines are 52% and46% respectively.

The increased production of electricity in the two countries haslargely come about through investments in thermal plants withcoal as the dominating source of energy. In 2003–2004, thermalpower plants accounted for 85% of the electricity generation inIndia (India Ministry of Power, 2010). The figure for the Philippinesis 56% (Del Callar, 2006). Whereas coal fired thermal plants used tobe highly pollutant, new cleaning technologies have reduced thelocal emissions to low and tolerable levels. Electricity generationby thermal plants is still not considered sustainable due to itseffect on climate change through emissions of large quantities ofcarbon dioxide.

ll rights reserved.

Hydropower plants have the advantage of relying on a renew-able energy source. Their emissions of climate gasses depend onlocation and design and may be insignificant. Their contributionto electricity generation has been relatively small. In India, hydro-power plants accounted for 14% of the electricity generation in2003–2004 (India Ministry of Power, 2010). The correspondingfigure for the Philippines was 19% (Del Callar, 2006). A main obsta-cle to the development of hydropower plants has been their effecton the environment and the local communities through riverdiversion, regulation of water flow, flooding and erosion and relo-cation of people. Plans for hydropower development often lead toprotests and social unrest. Many of the negative environmentaland social consequences may be prevented or compensated forthrough careful planning and design (IHA, 2004).

Smaller hydropower plants are generally not considered to beas intrusive to the environment and local society as large plants.Dams can be reduced in size or avoided altogether, thus avoidingmany of the problems of flooding, erosions and needs of relocationof people.

One aspect often overlooked is the high risk of accidents in hydro-power development projects. The construction industry is generallycharacterised by a high risk of accidents. According to ILO estimates,100,000 workers are killed annually on construction sites globally,which means that the fatality rates in construction is five timesthe general average among employees worldwide (Murie, 2007).

Historical data show that hydropower projects in particularhave been associated with many severe accidents. In Norway, for

Fig. 1. Overall analysis framework.

1942 U. Kjellén / Safety Science 50 (2012) 1941–1951

example, hydropower played an important role in the develop-ment of infrastructure during the last century, and high fatalityrates were tacitly tolerated by society. There were 23 fatalities inthe Tokke 1GW project, which was completed during the period1956–1972 (Skjold, 2006). Improved technology that reduced thenumber of workers at the risk exposed sharp end, and improvedsafety management routines decreased this number. The Ulla-Førre 2GW project was completed between 1974 and 1984 withthree fatalities. There were zero fatalities in the 182 MW Tyinproject in 2001–2004.

There is no reliable statistics on fatalities in hydropower pro-jects in India and the Philippines. Accident rates in the general con-struction industry are high, especially in India with the world’shighest accident rate among construction workers according toILO estimates as quoted by RICS (2009).

There are no reliable accident statistics for the Philippines. TheBureau of Labor and Employment Statistics in the Philippinesreported four fatalities in construction in 2003 which is likely fartoo low (BLES, 2008).

1.1. This paper

This paper reviews the experiences of an international hydro-power company (IHC), owned by the Norwegian government,and involved in developing and operating hydropower facilitiesin emerging markets primarily in joint ventures with indigenouspartners. The company primarily invests in smaller, run-of-riverplants. The company’s policy is to invest on commercial termsand ensure an adequate health, safety and environmental standardin the construction and operation of hydropower plants. This isaccomplished through the promotion of internationally-recognisedpower industry standards, and transfer of competence gainedthrough the development of hydropower projects in Norway.

In the start-up phase from 2002, the IHC took the role as finan-cial investor and had limited resources for technical and manage-rial follow-up of investment projects. Experiences with costoverruns, delays and poor safety performance in the company’sconstruction projects resulted in decisions to build up technicalexpertise and management systems for a closer follow-up of theinvestment projects. This build-up of resources took off in 2007.

The paper accounts for the experiences with two of IHC’s pro-jects from the start of construction until June 2010:

– A 196 MW Greenfield project in Himachal Pradesh in India,where construction started in 2005.

– A 75 MW Refurbishment project in Luzon in the Philippineswith start of construction in 2008.

In the first project, IHC had to intervene as a reaction to poorsafety and overall project performance when the project was threeyears into the construction phase. The second project illustrates aproactive position, where a new company policy on HSE manage-ment of construction projects was implemented in contract workand site follow-up. The paper addresses the following questions:

1. What has been the development in the overall safety performancein the two projects and to what extent can this be explained byIHC’s reactive and proactive interventions respectively?

2. Have there been any specific weaknesses in the safetymanagement during construction in the two projects indicatingshortcomings in IHC’s different approaches?

3. To what extent can the differences be explained by influencingfactors beyond IHC control?

4. What resources have been marshalled by the IHC in the twocases, and what implications does this have on the efficiencyof the two approaches?

5. Have there been any synergetic effects between safety andoverall project performance and to what extent can this beexplained by IHC interventions?

2. Materials and method

2.1. Overall analysis model

The analysis model underlying the study is shown in Fig. 1. It isassumed that the safety performance of each site is determined bythe site specific construction process, which also includes the pre-vailing safety practice. This study focuses on the resources mar-shalled by the IHC and the actual interventions to improve safetyat the sites and their effects on safety practices and performance.The results are interpreted in relationship to the moderatingeffects of project internal and external influencing factors.

2.2. Key data about the studied projects

The two projects were selected for this paper because they rep-resent two distinct different approaches in the IHC’s portfolio ofconstruction projects when it comes to the company’s follow-upof safety and overall performance.

Table 1 summarises some key facts about the projects.The influence of the different characteristics of the projects on

safety performance will be further elaborated in Section 3.2.

2.3. Points for intervention

2.3.1. Reactive approachIn the Indian project, design as well as contract structure were

decided before the IHC took on ownership. Before the intervention,the Indian project was followed up by the IHC through its repre-sentation on the JV Board and in the management committeeand through periodic progress reports. At the start, there was nosystematic safety performance reporting.

A technical expert was allocated to the project and visited thesite regularly. The IHC had also recruited an international experton tunnelling for secondment in the project.

It became evident in 2006–2007 to the IHC management that thesafety and overall performance of the Indian project were notacceptable. This was one of the reasons why the company in 2007decided to increase its resources for project follow-up in general.Technical staff was strengthened including the recruitment of aHSE manager at the end of 2007. Three fatalities in January 2008(one involving transportation and two off-duty fatalities due to anavalanche) further underlined the severity of the situation. An inter-nal report by the HSE manager early in 2008 documented the lack of

Table 1Basic facts about the projects.

Issue Indian project Philippine project

Type of project Greenfield RefurbishmentCapacity 192 MW 75 MWHead, water pressure exerted on turbines due to

distance between the water surface and theturbines

880 m 154 m

IHC ownership and position in Joint Venture 43% ownership with Indian company as majority owner 50% ownership, same as the JV partner, a companyfrom the Philippines

Construction period 2005–2011 2008–2011Nationality of project and contractor

management100% Indian, except for up to 10 international advisors(safety, tunnelling, general management) provided bythe IHC

International top management both for project andcontractors, Filipino middle management

Number of workers (peak) 4000 500Nationality of workers Ca 40% Indian, ca 60% Nepalese 100% FilipinoContract strategy Seven different contractors, about 20 major

subcontractors and multiple smaller subcontractors, allIndian

Two main international contractors for civil andelectro-mechanical respectively, ca 5 major localsubcontractors

Tunnels (new) 13,000 m 400 mTunnels (refurbished) None 120 mSite access roads 30,000 m 10,000 mSite elevation differences (intake – outlet) Ca 1000 m Ca 200 m

Table 2Overview of HSE interventions by the IHC.

Points forintervention:

Phase 1 (2008) Phase 2 (2009/Q1 + Q2) Phase 3 (2009/Q3 � 2010/Q2)

IHC audits Two, one in Q2and one in Q4

None Two, one in 2009/Q3 and one in2010/Q1

Organisationalchange

Not applicable – Secondment of an international HSE manager– Establishment of HSE Task force

– Secondment of internationalsafety expert responsible forinspections

– Secondment of internationaltransportation safety expert

Changes in HSEmanagementpractice

– The international HSE manager improved the routines for accident reporting andinvestigation, safety inspections, and housekeeping and introduced safety inductionand toolbox meetings

– IHC initiative to define safe work practice through implementation of job safetyanalysis

– Improved driver training andvehicle and road inspections

U. Kjellén / Safety Science 50 (2012) 1941–1951 1943

control of safety in the project, representing an ethical as well as rep-utation risk. It triggered work at a high managerial level in the IHC todevelop a strategy for project turn around. One critical issue was thecompany’s insufficient influence in the project due to a minorityposition. The strategy involved accident monitoring to determinepriorities and the use of audits to collect information about thecauses of the current situation and obstacles to improvements.The strategy also involved cooperation with the lender (IFC) tomobilise the necessary influence to accomplish change.

IHC’s intervention came about in three phases with increasingimpact. The presentation below focuses on safety and is summa-rised in Table 2.

Phase 1 was mainly characterised by information collection andanalysis on behalf of the IHC through audits and accident analysisin order to develop a strategy for change. The main changes wereimplemented in Phase 2, both through direct intervention by theIFC and through the secondment of international experts in theproject. The main focus in Phase 3 was to maintain the newlyachieved level of safety. Audits were actively used by the IFC inPhase 3 for this purpose.

The analysis in 2008 of the accident data and particularly thefatal accident statistics identified three major areas of concern:

– Internal transportation on site access roads typically involvingdrivers losing control of the vehicle and driving off the road.

– Slope instability along the site access roads resulting inconstruction workers being hit by boulders.

– Tunnel workers being hit by falling rock.

The analysis also showed that a general lack of a defined safetystandard and compliance in hazardous work was a major obstacleto safe work performance. The focus on safety inspections aimed atimproving the compliance culture in the company. Two accidentsin March 2009 resulting one fatality in tunnelling and two in trans-portation, triggered additional actions by the IHC. A team lead byIHC experts initiated job safety analyses (JSA) of the main subsur-face work at all work fronts. The JSAs defined the necessary mea-sures to prevent further severe accidents. The concrete measuresresulting from the activities in Table 2 are further described inSection 4.2.

The international HSE manager’s focus while in the project(Phase 2) was to upgrade the HSE management practices to inter-national standard. She was replaced by an Indian HSE managerwith international experience in Phase 3. A safety Task force withproject personnel and lead by an IHC manager was established latein Phase 2 to monitor safety work at the site on a regular basis. Oneof the duties of the Task force was to oversee that the JSAs werecompleted for all critical work and that the results were imple-mented in the construction work. Safety inspections were reorgan-ised and allocated additional resources in order to ensure at leastweekly coverage of all surface and subsurface work fronts withroutine inspections.

The IHC also introduced other organisational changes toimprove project overall performance on progress and cost. A sec-onded construction manager took a senior management positionin the organisation reporting to the manager for civil works, andthe international tunnelling experts oversaw the subsurface works

Fig. 2. The contracting process and integrated safety activities.

1944 U. Kjellén / Safety Science 50 (2012) 1941–1951

at the different fronts. A project steering committee with strongIHC representation was established for regular follow-up of theproject including monthly site visits. HSE has been an integral partof their follow-up.

To summarise, the additional IHC resources marshalled fromstart of intervention in January 2009 until end of the study period(June 2010) for follow-up include:

– Allocation of senior management resources from head office,approximately six man-months.

– Allocation of senior management resources from the company’soffice in Delhi, about six man-months.

– Secondment of HSE experts at site, one man year.– Secondment of international experts on construction manage-

ment and tunnelling, six man-years.

2.3.2. Proactive approachIn 2008, the IHC developed a new HSE policy including a policy

on follow-up of construction projects. In this part of the policy, thecompany had adopted an international best practice on safetymanagement of construction sites integrated in the contractingprocess, Fig. 2. This best practice is based primarily on oil andgas industry experience, where the Norwegian offshore industryhas taken an active part (OGP, 2010). One Norwegian companywith activities in different branches of industry had positiveexperiences of applying the same principles in its land-basedconstruction projects including hydropower projects.

The IHC’s policy on safety management in contracting was inthe process of being implemented when the Philippine projectwas in the tender preparation phase. Design had already beendecided in the JV with the IHC as technical operator. It had alsobeen decided to use international tender of the main civil and elec-tro-mechanical contracts. Two international contractors with agood reputation on HSE and overall performance had beenselected.

In accordance with the new IHC Policy, the contracts includedowner’s requirements regarding Contractors’ management of HSEand social issues at construction sites. These were documented ina separate specification that addressed:

– The responsibilities of the main contractors for HSE in own andsubcontractors’ construction work.

– The responsibilities of the civil contractor as principal contrac-tor for the coordination of HSE work and for emergency pre-paredness at site.

– Requirements to contractors’ HSE management systems andprocedures.

– Requirements to site induction and training, use of personalprotective equipment, minimum standard of safety at site, etc.

– Reporting requirements to the Owner.

Relevant IFC guidelines regarding HSE and social issues wereintegrated in the specification.

IHC lead a training session of project management on the prin-ciples of HSE management in the contracting process. Start-upactivities included evaluations of contractors’ HSE managementsystems and kick-off meetings to explain the Owner’s require-ments. These activities were conducted by the project manage-ment and the JV HSE manager, and IHC personnel were onlymarginally involved.

During construction (contract administration according toFig. 2), the project with support from the JV’s HSE manager con-ducted the follow-up of the contractors. The IHC’s follow-up ofthe project on safety had a routine character and included:

– One health and safety audit in Q2 2009.– Monthly follow-up on safety performance.– Regular contacts between JC’s and IHC’s HSE management in

case of severe incidents.– Coordination of emergency response activities in case of defined

emergency situations such as increased security threats,typhoons and flooding.

The IHC has seconded a international project managementexpert to the project as project manager. IHC has also been repre-sented on JV Board and Technical committee. The IHC resourcesput into follow-up on safety and performance of the Philippineproject are considered as normal for as project of this size andIHC ownership.

U. Kjellén / Safety Science 50 (2012) 1941–1951 1945

2.4. Data sources

The main source of data for this study comes from the regularwork with HSE in the project as available to the IHC. This mainlyincludes:

– Accident investigation reports and accident statistics reportedto the IHC. The accident statistics for the Indian project hasnot been reliable until after the intervention. Thus only acci-dents during Phase 2 and 3 have been included in the analysis.For the Philippine project, accident statistics for the wholestudy period has been included.

– Regular safety management reports.– Project progress reports.– Audit reports performed by the IHC and 3rd party. There have

been in all four audits by IHC of the Indian project and one ofthe Philippine project during the study period.

– Reports from site visits by IHC managers.

Interviews have been made for this study with key manage-ment personnel in the projects and with managers in the IHC withresponsibility for follow-up of the projects.

The analysis of external influencing factors has been based on aliterature review.

3. Factors influencing safety performance

3.1. External

3.1.1. National regulations and authority handlingBoth India and the Philippines have comprehensive occupa-

tional safety regulations (ILO, 2010). In India, the Ministry ofLabour and Employment has the overall responsibility for policieson occupational health and safety. The construction sector is regu-lated by the Building and Other Construction Workers Act from1996 and the associated Regulations from 2002. The regulationsare directed towards the employer, i.e. Contractor or the Clientwhen carrying out work directly (hired labour). The regulationsgive prescriptive requirements in such areas as fire protection,machine guarding, air quality, personal protective devices, electrichazards, lifting appliances, excavation and tunnelling work, scaf-folding, explosives handling, working hours, medical and welfarefacilities. There are also requirements to the establishment of ahealth and safety organisation at the construction site.

In the Philippines, the Department of Labor and Employment(DOLE) has the overall responsibility for safety at the workplaces.The Labour Code of the Philippines (Presidential Decree 442 0f1974) includes sections on labour protection. The requirementsto workplace safety are detailed in the Occupational Safety andHealth Standards from 1978 and amended in 1991. As in India,the requirements are largely prescriptive and cover all essentialareas of occupational health and safety. In the Philippines, everyconstruction project shall have a safety and health program, andit is the owner’s representative (project manager) that shall ensurethat it is established and implemented in accordance with thelegislation.

A crucial question is the extent to which the regulations areactually enforced. It is the author’s impression that there are signif-icant differences between India and the Philippines in enforcementwith direct effect on the two projects’ compliance with the regula-tions. Both countries suffer from infrequent and ineffective labourinspections due to a general lack of adequate resources (Deshing-kar, 2009; Hirose and Vitasa, 2007). In order to address thesituation, DOLE in the Philippines introduced a Labour StandardsEnforcement Framework (LSEF) in 2004. The joint venture

operating the project in Luzon has participated in a self-assess-ment scheme under this framework. It means that the project’ssafety and health program must be subject to regular third partyinspections and audits.

There is no similar scheme in India but for lifting tools, whichmust be subject to regular 3rd party inspections. None of the pro-jects in India and the Philippines have been subject to safetyinspections by government inspectors.

3.1.2. Non-governmental safety standardsBoth the project in India and the Philippines are subject to

financing by the International Finance Corporation (IFC). Thismeans that they must comply with IFC’s Policy and Performancestandards on Social and Environmental Sustainability (IFC, 2007).The IFC guidelines cover similar health and safety issues as thenational regulations in India and the Philippines but are not asdetailed. Compliance is checked by IFC through project reportingand regular independent reviews. These focus to a large extenton environmental and social issues. Relevant IFC standards havebeen reference documents in IHC’s audits of the projects, andnonconformities from these standards have been determined asespecially critical since they represent violations of the lendingagreement.

Of the occupational health and safety management system(OHSMS) standards that have been developed over the past30 years, the British Standard Institution’s specification OHSAS18001 has had an increasing impact (British Standards Institution,2007). It is compatible with the ISO 9000 series of quality manage-ment system standards and with the ISO 14001 standard onenvironmental management systems.

Both the IHC and the projects in Indian and the Philippines haveadapted it as the framework for their internal health and safetymanagement policies and systems. The Indian project has beenOHSAS 18001 certified by a 3rd party auditor.

Of the many standards and guidelines by trade associations onhealth and safety in construction, International Tunnelling Associ-ation’s Guidelines for good occupational health and safety prac-tices in tunnel construction have played the most significant role(ITA, 2008). It represents the construction industry’s best practiceon this subject and is a reference document in modern interna-tional tunnelling practice. Whereas the project in the Philippineshas followed these guidelines, implementation has been moreproblematic in the Indian project because of its use of a nationaltunnelling technology.

3.1.3. National economic wealth and cultureAs an investor in emerging markets, it has been a significant

concern to the IHC and its stakeholders whether the conditionsin the host countries will allow for the implementation of the com-pany’s safety policy of a high international standard. Two dimen-sions of special interest here are national differences in economicwealth and in risk-taking behaviour.

The purchasing power in India and the Philippines is about thesame but about 15 times lower than in Norway. The question ariseswhether an improvement in the safety standard has significantcost impact and whether the stakeholders in the host country arewilling to accept the increased expenditures.

Recent research has studied national differences in risk percep-tion and risk-taking behaviour at the workplace in relationship toHofstede’s cultural dimensions: power distance, individualism,masculinity, uncertainty avoidance and long-term orientation(Hofstede, 2004). In a study of six national workforce groups,including a sample of Filipino workers, only masculinity and powerdistance emerged as significant predictors of risk-taking behaviour(Mearns and Yule, 2009). The study also showed that perceivedsenior management commitment to safety is a more direct

1946 U. Kjellén / Safety Science 50 (2012) 1941–1951

predictor of safe behaviour than national culture. Workers in thePhilippines were engaged in significantly fewer risk-taking behav-iours, and showed low scores on Power Distance and a positiveperception of management commitment to safety. The low scoresfor Power Distance contradict Hofstede’s findings for Filipinonationals, which are even higher than those for India, indicatingthat Mearns’ and Yule’s findings apply to a corporate culture ratherthan a national culture. The implications are that a company man-agement that fosters a culture of low power distance also arereducing risk taking tendencies.

A study of construction workers in Pakistan confirms many ofthe findings by Mearns and Yule (Mohamed et al., 2009). An overallconclusion by Mohamed et al. was that workers’ intentional safebehaviour can be explained by their perception of corporate man-agement safety responsibility and of the risks in their work ratherthan by national cultural differences.

3.2. Internal

3.2.1. Nature of project and site layoutIn general, refurbishment projects are considered more accident

prone than Greenfield projects (Manu et al., 2010). This is due tothat hazards during refurbishment are more uncertain andcomplex. In the present cases, this situation is likely offset by thecomplexity of the geology in the Indian project, affecting safetyin tunnelling and on site roads. Tunnelling in this project has beenvery complex due to portions with very loose rock and large waterintakes, increasing the risk of falling rocks.

Site layout also differs considerably between the projects. TheIndian project is located between elevations about 2000 and3000 m above sea level and includes 30 km of internal access roadsto two separate intakes. The access roads have been carved out ofsteep mountains and there are significant risks of driving off theroads as well as risks of road erosion, falling boulders andlandslides. The roads are exposed to winter conditions with iceand snow and the area is avalanche prone.

Design in the Indian project did not mitigate the safety chal-lenges posed by the physical environment in the construction ofthe plant. The IHC promoted a solution involving the use of moderntechnique to build wide tunnels that could be used for transporta-tion as well. This solution would have reduced needs of usinghazardous mountain roads for transportation, but was not selected.

The IHC was involved in the design in the Philippine projectfrom the start, and it reflected the use of Norwegian hydropowercompetence. The site layout is more benign. The project is locatedbetween 500 and 750 m above sea level. There are about 10 km ofinternal access roads, but they are considerably less hazardousthan the mountain roads in the Indian project. The area is locatedin the path of seasonal tropical typhoons and is exposed to risks offlooding and land slide.

3.2.2. Management systems and practicesManagement support for safety is considered a key factor influ-

encing accident prevention practices and results (Cohen et al.,1975). The two projects differ considerably regarding constructionmanagement systems and practices, especially related to safety.The JV in the Indian project has an Indian majority owner thathasn’t shared the IHC’s ambitions regarding a safety standardbased on international best practice. This has been a significantarea of controversy during the construction period. In the Philip-pine project, both JV partners are aligned on HSE. In India, themajority owner’s position on the balance between safety and pro-duction has also been adopted by the local project management. Inthe Philippine project, top management of the project as well as ofthe main contractors have a shared understanding regarding theimportance of safety and accident prevention. These differences

are reflected in differences in compliance culture at the two sitesas well as in the preparedness to plan in advance and take thenecessary safety precautions before start of work.

3.2.3. Contracting philosophyA project’s contracting philosophy will affect safety performance,

because multiple contractors and subcontracting are considered tofragment the workforce and impede safety management at site(Manu et al., 2010). There are considerable differences betweenthe projects regarding contracting philosophy and use of thecontract in site follow-up.

Again the Indian project has a disadvantage with seven maincontractors, about 20 major subcontractors and multiple smallersubcontractors under the main contractors. This has rendered thefollow-up by the owners’ project team of safety at site more diffi-cult. The safety clauses of the contracts have been general in nat-ure, referring to the regulatory requirements but lacking projectspecific requirements, and incentives to comply with the contracthave been weak.

The Philippine project has adapted IHC’s contracting philosophywith two main contractors for civil and electro-mechanical worksrespectively with adequate qualifications within HSE. The IHC’srequirements to contractors’ safety management systems havebeen implemented in the contracts and followed up by the pro-ject’s site team and in audits. This has been one of the points forintervention based on lessons learnt from the Indian project.

3.2.4. Technology standardThere have been differences between the projects regarding the

use of modern construction equipment and the maintenance stan-dard of the equipment. Customs handling and requirements toadaptation to national standards has rendered the import of mod-ern construction machines more cumbersome in the project inIndia than in the Philippines. In India, there has been a mixtureof owner provided heavy equipment and equipment brought tosite by contractors and subcontractors. The owner provided equip-ment has been new and of high international standards, whereasthe contractor provided equipment usually has been second handand having low safety standards. In the Philippine project, themain contractors have bought new equipment, largely from recog-nised international suppliers. The technology and safety standardhas been high in international comparison.

In the Indian project, the owners’ project team has had limitedinfluence on the maintenance standard of the equipment providedby the contractors. In the Philippine project, the main contractorshave been responsible for the main equipment and they have exer-cised planned maintenance routines.

In the Indian project, there have been expectations to use labourintensive methods of work rather than modern, mechanised workmethods with few persons exposed to the hazards at the sharp end.Tunnelling is a typical example. There have been about 25 workersat each tunnel face in the Indian project, as compared to three inthe Philippine project.

All these differences have implications for the risk of accidents.Modern and manpower efficient construction methods will reducethe number of workers at the risk exposed sharp end and thusreduce the risk of accidents (Manu et al., 2010). This is especiallyevident in tunnelling, where the tunnel front constitutes a hazard-ous zone during certain parts of the production cycle. Poor mainte-nance will result in frequent breakdowns leading to increasedneeds of improvisations along with an increased risk of accidents(Saari, 1976/77).

3.2.5. Human resourcesEducation and experience among the construction workers dif-

fer considerably between the projects. Illiteracy has been an issue

U. Kjellén / Safety Science 50 (2012) 1941–1951 1947

in the Indian project. Whereas managers and supervisors in theIndian project have been Indian, about 60% of the workforce con-sisted of Nepalese migrant workers. The literacy rates in Indiaand Nepal are 61% and 49% respectively, as compared to 93% inthe Philippines (CIA, 2010).

The ratio between skilled and unskilled labourers has beenabout 20/80 in India and about 30/70 in the Philippines. The differ-ence between the two countries is less than expected from the dif-ferences in literacy rates, but the Philippine project has prioritizedlocal labour as part of the social responsibility program.

Families with children, where the man and wife formed a workunit, were common in the Indian project until the spring of 2009,when a strict ‘‘no children at site’’ policy was enforced. This tradi-tion with the family as a working unit did not exist in the Philip-pine project.

ig. 3. Development in the fatality rate in the Indian project by type of activity(N = 13).

Fig. 4. Distribution of fatalities in the Indian project by type of activity and event(N = 13).

4. Results

4.1. Safety performance

4.1.1. FatalitiesThe first fatal accident in the Indian project was reported in the

4th quarter of 2006. There have been in all 13 reported fatalitiesamong the JV’s and contractors’ employees in the Indian projectaccording to GRI’s definition of recordable fatality. There have beenan additional three fatalities involving construction workers, butthese have occurred outside of the project area and outside ofworking hours and thus have not met the GRI definition of ‘‘arisingout of or in the course of work’’ (GRI, 2006). There have been zerofatalities among JV and contractor employees in the Philippineproject.

Fig. 3 shows the distribution of the fatalities in the Indian pro-ject per quarter and by type of activity. Initially, accidents occur-ring in road work dominated. This has to do with the fact thatroad construction and maintenance work was most intensive dur-ing this period. From 2008, accidents during transportation havedominated. There have been no fatalities in the period after IHCintervention in Q1, 2009 and until the end of the study period(Q2, 2010).

The distribution of the fatalities in the Indian project by type ofactivity and type of accident is shown in Fig. 4. Almost half thefatalities can be attributed to vehicles driving off the mountainroads inside the site area. Falling rocks have caused five fatalities;three on roads during construction and two in tunnelling work.

Fig. 5. TRI-rate in the Indian project before and after IHC intervention and in thePhilippine project from start of construction. The TRI-rate is the total recordableinjuries per one million hours of work.

4.1.2. Total recordable injuriesThere was no reliable reporting of non-fatal accidents in the

Indian project before January 2009. In the analysis, the data fromthe first quarter of 2009 (first part of Phase 2 in IHC’s intervention)has been separated from the data for the rest of the study period. Alarge number of accidents were reported during Q1 of 2009, givingan unacceptably high TRI-rate of 49. During the second part ofPhase 2 and Phase 3 of IHC’s intervention (Q2, 2009–Q2, 2010),the TRI-rate has stabilized at an acceptably low average level of2. The average TRI-rate in the Philippine project since start of con-struction has been 6.

Fig. 5 indicates that less severe injuries go unreported in theIndian project after 2009/Q1, although the same reporting criteriaare used in both projects. The lost-time injuries completely domi-nate the TRI-rate in the Indian project after Q1 2009. This makes itunjustifiable to compare the safety performance of the two pro-jects based solely on the TRI-rates.

Fig. 6 shows the distribution of total recordable injuries in thetwo projects by type of accident.

F

It is interesting to note the types of accidents occurring in theIndian project after the TRI-rate has stabilized on a lower level:

– Falling rock has continued to cause injury after IHC interventionin spite of this being an area of priority together with motorvehicle accidents. Two of the falling rock injuries occurred intunnels and one on a site road.

– Almost half of the injuries related to ‘‘caught in/between objects’’.Two of the injuries were caused by unprotected moving machin-ery parts. One of them resulted in an amputation. The twomachinery-related injuries occurred in repair and cleaning work;these were auxiliary works not covered by the JSA.

– Two falling accidents after IHC intervention were caused by apoor work platform and inadequate access respectively.

Fig. 6. The total number of recordable injuries per one million hours of work by type of accident in the Indian and Philippine projects.

1948 U. Kjellén / Safety Science 50 (2012) 1941–1951

– An inadequate rigging arrangement resulted in a severe injury,when the victim was hit by a ruptured sling (struck by movingobject).

– There were no reported injuries due to motor vehicle accidentsin the period.

The injuries in the Philippine project have generally been lesssevere. Accidents caused by the victim’s or co-worker’s movementsor manual handling (strike against, struck by, and caught in/be-tween) have dominated the picture. There have been three fallingobject accidents; the most severe being caused by a falling tree.

4.2. Physical safety standard

4.2.1. Indian projectIHC personnel observed significant violations of regulatory

requirements and recognised construction practice in the Indianproject during the early site visits and audits in 2008. These viola-tions included children engaged in project work as part of familyunits, lack of proper arrangements for safe work (e.g. work plat-forms, guardrails, lifting equipment), inadequate use of personalprotective equipment and poor housekeeping.

Through the audits in 2008–2010, developments in the safetystandard could be registered in a systematic way.

The work to protect construction roads by roadside barriers hadstarted already in 2008 before IHC’s intervention, but work wasintensified, and the quality and coverage of road side barriersincreased as a result of IHC intervention. At the end of the studyperiod, all site roads representing a hazard due to vehicles drivingoff the road were covered by such barriers. A transportation man-agement system has been introduced during the study period,including the introduction of traffic safety rules (speed limits),training of drivers and control of driver qualifications, inspectionsof vehicle road worthiness and road standard, and control of accessto construction roads during severe weather conditions. The risk offalling rocks has been mapped on the most exposed site road, andmeasures have been implemented to secure the road. There havebeen no recorded transportation related injuries during Q2 2009–Q2 2010; indicating that the different measures have been efficientin reducing the risk in this activity.

The second focus area in the IHC’s efforts to reduce the risk ofaccidents has been tunnelling. There have been three major areasof concern: the risk of falling rock, poor ventilation, and inadequateemergency escape equipment and plans. During the study period, asystem for securing of rock based on international standard hasbeen introduced. Ventilation has been improved step-by-step,

whereas work to improve emergency evacuation has been laggingbehind. The two falling rock injuries in tunnelling after IHC inter-vention indicate that this work has only been partially successful.

There have been no issues with children engaged in projectwork since Q1 2009.

There have been considerable improvements in the availabilityand use of PPE in tunnel work as well as surface work including re-pair work on construction roads. During the later audits and siteinspections, use of PPE was next to consistent.

The housekeeping standard has improved considerably duringthe study period, although setbacks have been observed. A house-keeping award system was introduced in Q2 2009 and has had po-sitive effects. The safety standard of access and work platforms hasvaried, and two fall injuries after IHC intervention can be attrib-uted to deficiencies in this respect.

The quality of contractor supplied equipment such as liftingequipment and cement mixers has been a recurring problem, andno significant improvements have been observed. At least threeinjuries after IHC intervention can be attributed to contractor sup-plied equipment of a poor standard, two of which with severeconsequences.

4.2.2. Philippine projectAudits and inspections by IHC personnel have shown that the

safety standard in the Philippine project has been adequate ininternational comparison.

The standard of vehicles and equipment has differed consider-able from the Indian project. Besides from some subcontractorsupplied vehicles, these have been new equipment. The maincontractors have been responsible for maintenance of most of theequipment and this has been done in accordance with goodinternational practice.

Another notable difference is in the area of transportationsafety. The principal contractor has implemented a rigorous trans-portation safety management system according to internationalbest practice. There have been some transportation related inci-dents at the site but no injuries or fatalities.

The Philippine project has applied international best practice intunnelling including internationally accepted methods for rocksupport.

The audit identified nonconformities particularly in the safetystandard of some minor work performed by small contractorsreporting directly to the owner’s organisation at site, indicatingthat these contractors had fallen outside of the main contractingprocess.

ig. 7. Developments in the RUO-rate for the projects in India and the Philippines.he RUO-rate is defined as the number of reported unwanted occurrences perorker and year.

U. Kjellén / Safety Science 50 (2012) 1941–1951 1949

4.3. Safety management systems and practices

4.3.1. Documented health and safety management systemBoth projects had documented stand-alone health and safety

management systems. The Indian project’s system had been OH-SAS 18001 certified in 2008 and this certification was renewed in2010. Audit results showed that it was not a living system in useby the JV’s site management. JV management largely related tothe project’s common procedure manual. This was successivelyupdated with new health and safety procedures in key areas inPhase 2 and 3 of IHC’s intervention and successive audit resultsshowed that compliance was improving. These procedures coveredaccident investigations, safety inspections, toolbox meetings, traf-fic management, etc. In addition, some of the main contractorshad their own health and safety management systems.

In the Philippine project, both the JV and the main civil andelectro-mechanical contractors had their own documented healthand safety management systems. The latter also applied to subcon-tractors. The JV’s system was compliant with OHSAS 18001,although it had not been certified, and with IFC standards as arequirement from the lender. Contractors’ systems should meetthe owner’s comprehensive contractual requirements to healthand safety management, and the audit showed that they werelargely compliant.

One critical issue was the coordination of health and safetywork including emergency response planning at site. In the Indianproject, the JV’s site organisation had taken this responsibility. Itwas a difficult task due to the multitude of contractors and subcon-tractors. In the Philippine project, this was the responsibility of thecivil contractor. Audit results in Q2 of 2009 showed that thisresponsibility was not fully understood, and corrective actionswere implemented as a result.

4.3.2. Safety practiceIn the Indian project, a number of improvements of the safety

practices were implemented during the two first quarters of 2009:

– The quality of the accident investigations improved consider-ably through participation of safety personnel with adequateexpertise. There was an increased willingness to address techni-cal and procedural nonconformities and needs of efficientpreventive measures.

– Job Safety Analysis was initiated by IHC as a proactive tool forthe planning of work and the identification of the necessarysafety precautions. All main activities of a critical nature havesubsequently been covered by JSA, but accident investigationshave shown that support activities such as cleaning of equip-ment have not been adequately covered. There have also beendeficiencies in implementation of results.

– The safety inspection routines have improved considerablythrough the mobilisation of dedicated personnel to this task.In practice, all work fronts have been covered at least thoughweekly inspections. Follow-up of results has lagged behinddue to inadequate line management involvement.

– Induction training of construction workers never materialisedgenerally at the site. This was compensated for through weeklytoolbox meetings with all construction personnel conducted bythe owner’s safety engineers.

– RUO reporting was introduced as a focused safety activity in allIHC’s construction projects and operating plants from January2009. Fig. 6 shows the development for the Indian and Philippineprojects. The Indian project started to report RUOs in July 2009,but the results never reached the IHC’s target (RUO-rate >1).

In the Philippine, the two main contractors had the responsibil-ity for the safety activities at site as specified in the contractual

FTw

requirements. IHC audits and other types of follow-up workshowed that the contractors were largely compliant. All the basicsafety practices such as safety inspections, accident reporting andinvestigations safety induction, and use of job safety analysis. func-tioned adequately. The RUO-reporting has been satisfactory, Fig. 7.

4.4. Overall performance

Initially, the Indian project experienced significant delays andcost overruns. The IHC’s intervention aimed at improving projectperformance on cost, progress and quality as well as HSE. The pro-ject’s improved safety performance was paralleled by significantimprovements in predictability regarding budget and schedule.

The Philippine project has delivered on cost but has experi-enced delays due to technical challenges related to the refurbish-ment of existing facilities.

5. Discussion

5.1. Development in the safety performance in the two projects

The Indian project has shown a satisfactory development in thefatality rate during the last part of the study period after IHC’sintervention, as compared to earlier phases of the interventionand the time before the intervention took place. Also the TRI-rateshows an improvement. In the last phase, the project has deliveredsafety performance results comparable to a high internationalstandard. Although not statistically proven, the data seen in con-junction make it highly likely that IHC’s intervention has had a sig-nificant positive effect.

The development of the fatality rate can partly be explained bya reduction in accident prone activities such as road construction.Civil work including tunnel construction and transportation haveremained at a high activity level throughout the study period,and the directed mitigating measures at these activities in particu-lar likely have made considerable contributions to the positiveresults.

Due to the absence of reliable non-fatal accident data during theproject period before IHC intervention, only data from 2009/Q1until the end of the study period were included in the analysis.The data for Q1 of 2009 is regarded as exceptional, since it demon-strated the safety performance results when reporting reliability isconsidered to be adequate but before the effects of IHC interven-tion on the accident risk have taken effect. It is a short periodduring the winter season and is not fully representative for the

1950 U. Kjellén / Safety Science 50 (2012) 1941–1951

conditions before IHC intervention. It gives together with fatalaccident statistics, however, an indication of the safety level duringthat period.

There is a considerable reduction in the TRI-rate from 2009//Q1to the last part of the study period (2009/Q2–2010/Q2). The acci-dents risks that continued to produce injuries after the IHC inter-vention had taken effect show that all measures have not beenfully efficient. Rock fall was still a problem in spite of the effortsbeing taken, demonstrating both inadequacies in implementationand the challenging geological environment in which the projectwas executed. Accidents due to poor work platforms and inade-quate access indicate remaining problems related to compliance.Two severe accidents due to moving machinery parts demon-strated shortcomings in the implementation of JSA in auxiliarywork. The efforts to reduce the risk in transportation paid offthrough satisfactory safety performance results.

It is evident from the breakdown of the recordable injuries byconsequence that the reporting pattern has changed; the lesssevere medical treatment and restricted workday injuries beingabsent in the materials in the last part of the study period. Evenif we double the TRI-rate due to assumed missing data, the resultsare still satisfactory in international comparison.

Besides from statistical fluctuations, the TRI-rate for the Philip-pine project shows no significant development in any directionduring the study period. The project has delivered acceptablesafety performance results during the whole period.

It is not possible to separate the effects of IHC’s new policy onsafety in construction on these results, since it has been an integralpart of the company’s general intervention on project managementin the Philippine project.

5.2. Safety management in the two projects

The present study has shown that there were considerableimprovements in the Indian project’s safety management systemsand practices as a result of IHC’s intervention; the study alsorevealed significant remaining deficiencies resulting in severe inju-ries. The safety management in the Indian project needs to be eval-uated with reference to an Indian context. The Indian constructionindustry is renowned for a poor HSE standard and high incidentrates. Very few Indian construction companies have a fully devel-oped safety management system in place. The Indian project didnot differ negatively in any significant way from this standard evenbefore IHC’s intervention. The identified deficiencies may beexplained by differences in expectations between the Indian pro-ject organisation and the IHC resulting in a lack of commitmentand endurance when it comes to meeting IHC’s expectations.

The Philippine project has delivered according to expectationswhen it comes to safety management. The civil contractor had,however, some initial difficulties in comprehending the full rangeof its responsibilities as principal contractor, indicating needs ofstrengthening the kick-off activities at construction start-up. Thefact that small contractors reporting directly to the owner fell out-side the safety management policy adapted by the project for maincontractors shows that the policy needs some amendment.

An interesting aspect is the significance of the OHSAS 18001standard as to the safety management in the projects. The scientificevidence on the efficiency of OHSMS standards in improving safetyperformance is inconclusive, largely due to the inadequate qualityof the evaluation studies on the subject (Robson et al., 2007). TheIHC early audits of the Indian project have revealed significantdeviations from the OHSAS 18001 standard in spite of the JV’s siteorganisation being certified. This finding indicates weaknesses inimplementation rather than in the standard itself and shows thatcertification is no guarantee of compliance. Safety performanceresults show, however, that a positive development partly can be

attributed to improvements in some very basic safety practices.The fact that the standard reference for these practices has beenOHSAS 18001, since India lacks adequate regulations for internalcontrol of safety, indicates that it has played an important role inensuring an adequate standard on safety management.

5.3. The significance of influencing factors

It is evident from the review in Section 3 that the conditions inthe two projects are radically different when it comes to influenc-ing factors. For most external as well as internal influencing fac-tors, the Philippines project had an advantage regarding the basicconditions required to achieve a good safety performance.

The pure size of the projects differs by about a factor of ten. Thephysical environment in the Indian projects is considerably morechallenging from a safety point of view. These factors alone mayexplain a significant part of the differences in fatality ratesbetween the projects. The Philippine project also had the advanta-ges of a considerably less complex contracting setup, more modernequipment and qualified personnel; all factors contributing toimproved safety.

A notable difference between the two projects has to do withthe compliance culture. The results cannot be explained bynational cultural differences. The Philippines score even higherthan India on Hofstede’s Power Distance and Masculinity indices,which should make the country’s construction workers and man-agers more prone to set aside safety rules and take risks (Hofstede,2004). A likely explanation is differences in corporate culture in thetwo projects as expressed by corporate and project management’scommitment to safety and to comply with safety rules. The factthat the Philippine project, as opposed to the Indian project, hadselected international main contractors with a good record onsafety performance in the past, may also explain the difference incompliance behaviour. The results coincide with the findings ofMearns and Yule (2009) and Mohamed et al. (2009).

A relatively high degree of illiteracy among the workers in theIndian project has been a concern. As a result of IHC’s intervention,the Indian project implemented oral presentations to the workersof safety rules in connection with safety inductions and toolboxmeetings more strict enforcement. These measures have likelyhad effect on the compliance culture, although quantitative evi-dence is largely missing. Illiteracy has not been an issue in the Phil-ippine project.

Some of the issues have been project specific; others are appli-cable to construction projects in the two countries in general. Itremains to be demonstrated that IHC’s proactive policy in safetymanagement in construction projects is fully applicable in anIndian context and will give the same positive results as in thePhilippine project.

5.4. IHC resources

The study shows that IHC needed to marshal considerable man-agement resources and personnel to secure the necessary improve-ments in safety performance in the Indian project. The good resultsin the Philippine project were possible to accomplish within anordinary budget frame for project follow-up. These differences can-not be explained fully by the reactive approach being much lessefficient. External and internal influencing factors have played amoderating role and have largely given the proactive approach inthe Philippine project an advantage as compared to the reactiveapproach in the Indian project. It is likely that an application of aproactive approach in an Indian context would be more resourcedemanding.

U. Kjellén / Safety Science 50 (2012) 1941–1951 1951

5.5. Synergy effects between safety and overall project performance

The Indian project has shown a considerable improvement inpredictability regarding cost and schedule after the IHC interven-tion, whereas the Philippine project largely has performed wellon cost and schedule throughout the construction period. The pro-jects’ results on safety and overall performance seen in conjunctionindicate synergetic effects between safety and general projectmanagement that would not be possible to accomplish if the inter-vention focused on one of the areas alone (Kjellén et al., 1997).

5.6. The significance of the findings

The paper presents two case studies representing two differentstrategies in an international hydropower company’s interventionto ensure an acceptable safety standard in construction projectsin emerging markets. The study is mainly based on regular reportsfrom the two projects, including reports on accident statistics andowner audits, and there has been no systematic data collectionthroughout the study period in the two projects specifically for re-search purposes. The shortcomings are evident, notably concerningnon-fatal accident statistics in the Indian project before IHCintervention.

The two projects represent two different intervention strategiesby the IHC. Due to the nature of the study, it has not been possibleto select two projects that have been reasonably similar regardinginfluencing factors in order to isolate the effects of the interventions.Rather, the findings from the study must be judged in light of thestrength in the differences in safety performance between the twoprojects and how the Indian project has developed over time.

6. Conclusions

The study has demonstrated that a proactive approach, involv-ing use of the contracting process to accomplish the necessary pre-conditions for a safe execution of a construction project, hasdefinite advantages. It has the potential of delivering satisfactorysafety results at moderate costs for follow-up. A disadvantage withthe reactive approach is poor safety performance in the period be-fore the corrective actions have been implemented and have takenfull effect. It is also resource demanding and requires endurance.

In both studied projects, the IHC’s interventions have includedimprovements in general project management with safety as animportant component, and the results show the adequacy of thisholistic approach.

Some of the differences between the projects may be explainedby project specific conditions and differences in prevailing normsregarding risk acceptance in the two countries. It is likely that,due to cultural differences, the application of a proactive approachwill be more resource-demanding in India, as compared to thePhilippines.

Acknowledgements

The author thanks Partha Saha, Rodolfo Azanza, and Jarl Kos-berg, SN Power, and Nomer Reynaldo, SN Aboitiz for valuableinput.

References

British Standards Institution, 2007. Occupational Health and Safety ManagementSystems – Requirements. BSI Standard OHSAS 18001:2007, London.

BLES, 2008. Philippine Industry Yearbook of Labor Statistics, Philippine Bureau ofLabor Statistics.

CEA, 1999–2009. Monthly and Annual Reports. Central Electricity Authority, NewDelhi.

CIA, 2010. The World Factbook, Central Intelligence Agency, Washington. <http://www.cia.gov/library/publications/the-world-factbook/>.

Cohen, A., Smith, M., Cohen, H., 1975. Safety Program Practices in High vs. LowAccident Rate Companies. NIOSH, Cincinnati.

Del Callar, C.C., 2006. Sustainable Power Generation in a Restructured PowerIndustry. National Power Corporation, Quezon City.

Deshingkar, P., 2009. Extending Labour Inspections to the Informal Sector andAgriculture. Chronic Poverty Research Centre, Working Paper No. 154, London.

GRI, 2006. Sustainability Reporting Guidelines. Global Reporting Initiative, Version3.0, Amsterdam.

Hirose, K., Vitasa, B.C., 2007. Implementation of the Labour Standards EnforcementFramework in the Philippines, Working Paper 14. ILO, Manila.

Hofstede, G., 2004. Cultures and Organizations: Software of the Mind. McGraw-Hill,New York, New York.

IHA, 2004. Sustainability Guidelines. International Hydropower Association,London.

IFC, 2007. Environmental, Health, and Safety (EHS) Guidelines. IFC Environment andSocial Development Department, Washington.

IHS Global Insight, 2010. World Economic Service. Lexington, MA.ILO, 2010. NATLEX Database of National Labor, Social Security and Related Human

Rights Legislation. ILO’s International Standards Department, Geneva. <http://www.ilo.org/dyn/natlex/natlex_browse.home>.

India Ministry of Power, 2010. Economic Survey 2006–2006.ITA, 2008. Guidelines for Good Occupational Health and Safety Practice in Tunnel

Construction. ITA Report No. 001, Lausanne.Kjellén, U., Boe, K., Hagen, H.L., 1997. Economic effects of implementing internal

control of health, safety and environment: a retrospective case study of analuminium plant. Safety Science 27, 99–114.

Manu, P., Ankrah, N., Proverbs, D., Suresh, S., 2010. An approach for determining theextent of contribution of construction project features to accident causation.Safety Science 48, 687–692.

Mearns, K., Yule, St., 2009. The role of a national culture in determining safetyperformance: challenge for the global oil and gas industry. Safety Science 47,777–785.

Mohamed, S., Ali, T.H., Tam, W.Y.V., 2009. National culture and safe work behaviourof construction workers in Pakistan. Safety Science 47, 29–35.

Murie, F., 2007. Building safety – an international perspective. International Journalof Occupational and Environmental Health 13, 5–11.

OGP, 2010. HSE-Management Guidelines for Working Together in a ContractEnvironment. International Association of Oil and Gas Producers. Report No.423, London.

RICS, 2009. The Labour Shortage is for Real: Education and Professional Skills. RoyalInstitution of Chartered Surveyors, London.

Robson, L.S., Clark, J.A., Kullen, K., Bielecky, A., Severin, C., Bigelow, P.L., Irvin, E.,Culyer, A., Mahood, Q., 2007. The effectiveness of occupational health and safetymanagement system interventions: a systematic review. Safety Science 45,329–353.

Saari, J., 1976/77. Characteristics of tasks associated with the occurrence ofaccidents. Journal of Occupational Accidents 1, 273–279.

Skjold, D.O., 2006. Statens Kraft. Universitetsforlaget, Oslo.