Housing Reconstruction - nicee.org

17HOUSING RECONSTRUCTION

Relocation VillagesObservations■ To handle rebuilding of entire villages at new

sites, the government needed to use contractors.Local artisans were not available in the numbersrequired. Scale of the work was too large to bemanaged by the beneficiaries themselves in theinitially-assumed implementation period of threeyears, particularly given the traumatized psycho-logical state the GOM perceived in the villagers.

■ While some evaluation studies suggest that theprogram design for the relocation villagesdampened the initiative of the beneficiaries, incontrast to the repair and strengthening programwhere beneficiaries were actively involved in therebuilding decisions, a comprehensive survey ofevery beneficiary in the relocation villages founda high level of satisfaction with the new houses.

■ Extensive use of concrete technology in therebuilding presented problems related to thequality of construction. It would have been betterto use alternative technologies, even at the cost ofprolonging construction and extending theimplementation period.

DiscussionThe most affected villages, practically reduced torubble in the earthquake, were considered forreconstruction at new sites (relocation). Relocationvillages were those in which more than 70 percentof the houses were either destroyed or substantiallydamaged according to the IAEE Damage Categories4 and 5 (IAEE, 1986), and those villages located onsoft soil (so-called “black cotton soil”) of more than2 m (6.5 ft.) depth. Villagers believed that sincemost of the earthquake fatalities occurred in thesevillages, they had become cremation sites and burialgrounds and thus were uninhabitable for psycho-logical and religious reasons. Arguments for reloca-tion also included the fact that because there was somuch debris in the totally destroyed villages, it wasnot economically feasible to clear it out and rebuildat the same sites. There was also a fear that blackcotton (expansive) soils would make villages

The rebuilding project represented a complex efforton many fronts. Not only did the GOM create aseparate management structure for the project, but italso developed three alternative approaches forrebuilding.

1. Relocation was selected as most appropriate forthe most severely damaged villages. Constructionwas managed by contractors under the supervi-sion of the PMU or by NGOs and donor agenciesthat adopted certain villages.

2. Twenty-two other villages ultimately fell into thehybrid approach-–a new village, usually built ona new site, with construction managed by thePMU or supervised by owners and NGOs.

3. The reconstruction, repair and strengtheningprogram was developed to rehabilitate moder-ately damaged houses in-situ. Under this pro-gram, beneficiaries were given the options ofbuilding a new safe room onto their houses orof strengthening the existing house with earth-quake-resistant technology. Owners managedthe repairs with technical support from thegovernment.

Immediately after the earthquake, the GOM beganconsidering options for rebuilding on the massivescale that was required. The GOM developed itshousing rebuilding policy based on the extent ofdamage and the earthquake fatalities. Political,psychological, religious, technical, and pragmaticreasons all played a role in the government’sdecisions concerning available options.

There were three general categories of damage:housing in completely destroyed villages, severelydamaged housing to be replaced in-situ, andmoderately damaged housing to be repaired orrebuilt in-situ. Packages of financial assistance weredeveloped according to these three general damagecategories. These categories also corresponded tothe three alternative approaches for rebuilding.

Housing Reconstruction

18 LESSONS LEARNED OVER TIME

vulnerable in future earthquakes. The expansivenature of black cotton soil not only necessitated theconstruction of deep foundations (i.e., either deepstrip foundations or piles), it also had substantialcost implications. In addition, popular sentimentconsidered relocation an opportunity to provideearthquake victims with “well-planned and neatlylaid out new villages at new sites without anysegregated compartments for different castes andcommunities” (GOM, 1993). Villagers, reinforced byprominent social science institutions, pleaded forrelocation. The government responded politically tosuch strong sentiment by agreeing to the relocation.More than 27,000 houses were ultimately relocatedin over 52 villages.

From the first days after the earthquake, governmentofficials recognized the dilemma inherent in thedecision to relocate entire villages, acknowledgingthat it was not feasible to abandon the area wherepeople had lived for generations, had establishedsocial relationships, and were close to their agricul-tural lands. The scientific basis for relocation wasalso not strong, since the sites were only to bemoved a few kilometers at most. On the other hand,the villagers were not psychologically prepared tostay in their destroyed villages, and consequentlythey put enormous pressure on the GOM to movethem. Some of the villages absolutely refused toconsider rebuilding in-situ, and although the GOMattempted to convince villagers that rebuilding in-situ would be safe, they were ultimately unsuccess-ful. The GOM felt obliged to be responsive to thedemands of the villagers. It is important to empha-size that the demand for relocation was very pro-nounced, and the GOM did not believe it had thetime or resources necessary to mount a majoreducation campaign to convince villagers otherwise.

As the GOM evaluated the option of a more tailor-made approach to rebuilding houses in the severelydamaged villages, they decided that given the scaleof rebuilding required, it would be impossible torebuild in a timely fashion if each house weretreated as a separate project (find the space withinthe house, clear debris, modify each design, findindividual builders, etc.). There were significantadministrative issues in thinking through costsassociated with a less standardized package. (Couldexisting foundations be used? If not, how wouldnew foundation costs be factored in in a uniform

manner? How would debris removal costs behandled in a uniform fashion? Etc.) These significantadministrative issues were important to the GOM indesigning a program that would have a high prob-ability of success in terms of the number of unitsrebuilt in a timely fashion, improvements in thestandard of living, and enhancements in earthquake-resistant construction (Vatsa, 1999) (Figure 8).

The new villages were located close to the old.There were three house types available under thisprogram, which varied in size. Type A was 250 sq.ft. (this was the core house that most peoplereceived as grant assistance), Type B was 450 sq. ft.,and Type C was 750 sq. ft. The size of the house thebeneficiaries had before the earthquake determinedwhich size they received. (Details are available inthe “Earthquake Rehabilitation Policy of the Govern-ment of Maharashtra,” GOM, 1994a.) The sitesselected for resettlement were acquired eitherthrough voluntary transfer to the GOM by thelandowners or land acquisition was carried outunder relevant law. Ownership rights to the houseplots in the original villages were surrendered to theGOM upon acceptance of the new houses. Whilesome of the NGOs involved in various relocationvillages later reported that the resettlement awayfrom the original land was inevitable and unavoid-able, others thought that the siting of new settle-ments at original places would have made better useof the old plots and materials (CSSS, 1998).

Immediately after the earthquake there was anoutpouring of help from various NGOs and chari-ties. Since the World Bank-assisted program was notofficially launched until June 1994, when credit fromthe World Bank was received, many organizationsbegan rebuilding villages at new sites before thisprogram was in place. The donor agencies madedecisions regarding construction materials andtechnologies that did not necessarily conform withthe guidelines later developed by the GOM. OtherNGOs committed to rebuilding housing, but waiteduntil the GOM program was in place to make theirunits conform to those that were built under thegovernment program.

Although an initial principle of the program was toinvolve the beneficiaries directly in the rebuilding,the GOM decided to let contracts for the construc-tion of entire villages because of the scale of the


Figure 8 Category A (relocation) and Category B (hybrid) villages in the Latur and Osmanabad districts.

relocation and the limited number of local artisansavailable for construction. Different contractorswere selected for each of the relocation villagesand were then responsible for all the constructionin the village. These contracts were managed bythe Project Management Unit (PMU) created for thisproject. Engineering consultants and rural resettle-ment planners carried out village planning anddesign and supervised the contractors who did theconstruction (Figure 9).

A total of 27,919 houses were constructed in therelocation villages. Construction of the majority ofhouses (19,513) was managed by the GOM. Donoragencies and NGOs constructed the remaining8,406 houses. The NGOs and donor agenciesmanaged construction in the smaller villages, andbecause they were not waiting for financing fromthe World Bank they were able to proceed more

quickly. Once they finished their own programs, theGOM asked them to continue construction at govern-ment cost. The GOM gave the NGOs certain conces-sions to encourage their participation, includingexemptions from sales and excise taxes, free waterand electricity during construction, and the land.

One unforeseen consequence of relocation was anaggravation of the problem of the drinking watersupply. According to some of the NGOs working inthe relocation villages, water supply in this drought-prone region has always been problematic. However,moving the villages 2 to 4 km (1.2 to 2.5 mi.) increasedthe problem for several reasons: the village wasresettled away from the existing water source, theinferior construction quality of the drinking waterpipeline, and construction of the drinking watersystems was delayed (CSSS, 1998).


• Secondly, the GOM had to appoint the designconsultants, who were responsible for preparingthe designs and the tender documents for 52villages. Local competitive bidding was used forthis purpose and the World Bank procurementprocedures, new to the PMU officials, had to befollowed. As of June 1995, 62 percent of thecontracts had been awarded and another 17percent were in the tendering stage.

• To plan the villages (village layouts) and to allotplots to individual beneficiaries, the PMU hiredcommunity participation consultants to workwith the villagers, the design consultants, andthe district administration.

• After the contracts were awarded, the designconsultants prepared the house designs (generallythree typical designs for each village), which hadto be approved by the project’s chief engineer.Most consultants hired rural architects and/or ruralresettlement planners to assist them in the design.

Construction ProcessGiven the issues of land ownership, home owner-ship, beneficiary preferences, and contractor man-agement, the construction process for houses in therelocation villages was complex. The key steps areoutlined below. The MEERP was launched on June30, 1994, nine months after the earthquake, and asnoted above, although the government-sponsoredrebuilding did not begin until then, some donoragencies had begun construction earlier.

• First, the GOM had to acquire land for the plotsin all 52 villages. As an example, the total areaacquired for the construction of 2,551 houses inthe Killari village, the largest village and the onemost affected by the earthquake, was over 203hectares (502 acres) (GOM, 1998c). It took theGOM over a year to complete the land acquisi-tion process. By June 1995, in the Latur andOsmanabad districts 94 percent and 87 percentof land was acquired.

Figure 9 Manufacturing of concrete blocks by a contractor using a block-making machinein a relocation village.


In general, the design consultants appointed werefrom the largest and most reputable engineeringconsulting firms in Maharashtra and India.

• The final step in the process was the construc-tion of the houses. The design consultants hiredthe contractors to carry out the construction.Due to a lack of interest by local contractorsfrom the affected area and even from Mumbai,the large majority of the contractors came fromthe neighboring state of Andhra Pradesh. Theconstruction was supervised by the consultantsand the PMU engineering staff.

Some of the reasons for the slow initial progress ofthe PMU-constructed housing are documented inprogress reports from the first two years of theMEERP implementation. These reasons include anacute shortage of water in the general drought-pronearea and erratic power supply; shortage of goodquality construction materials like sand/brick/hollowblocks; revisions in the beneficiary lists due toincreases in the number of claimants; inability of thecontractors to provide adequate manpower andmachinery; delay in land acquisition due to bifurca-tion problems (some castes and communities wantedto be separate) and demand for undue compensa-tion; an acute shortage of skilled construction work-ers; work stoppage by villagers for undue demands;and delays by donors in turning over sites in villagesconstructed jointly by the PMU and donor agencies.

New Village DesignFrom the earliest days immediately after the earth-quake “modern and neat settlements” were called forby village leaders, prospective beneficiaries and theGOM’s own policy document (GOM, 1993). Throughthe community participation process, villages weredesigned to meet the wishes of the beneficiaries.Later, there was criticism of the “grid” layout (Figure10). However, it did expedite construction, improvecirculation patterns, and reduce congestion. In fact,most of the donor agencies also used grid layouts inthe villages they constructed. Better infrastructurefacilities were provided. In some villages, attemptswere made to maintain traditional character andlinkages. In addition, the land area in the newvillages was significantly larger, allowing space forfurther extensions to individual homes. The majorityof the traditional rural houses in the area are onestory and hence only a horizontal building extensionis possible. In a large survey in which 23,498 benefi-ciaries were interviewed, and which was conductedas one of the evaluation components of the rebuild-ing project by the Center of Studies in Social Science(CSSS), 77 percent of the respondents were of theopinion that the design of the new village was betterthan the old. Only 10 percent believed that theoverall design of the old village was more convenientthan the new. This same study surveyed 13 NGOsthat had been involved in the relocation villages.Many of these NGOs argued that the basic design for

Figure 10 A grid-likelayout of a relocationvillage, Latur district


the new houses was not suitable for agriculturalcommunities.

The community participation consultants hired bythe PMU developed campaigns to disseminateinformation and elicit peoples’ views on the MEERP.This was the first time that the GOM had usedcommunity participation consultants in a program,and they were given varying degrees of support. Insome instances their role was welcomed, while inothers it was viewed as “unnecessarily interfering”(TISS, 1997). They were brought in late, after anumber of critical policy decisions had been made.In general their role included familiarizing govern-ment officials with the process of communityparticipation, forming the village-level committees(VLCs), capacity building through various trainingprograms, and information dissemination anddocumentation (TISS, 1997).

To help with information dissemination, the commu-nity participation consultants along with three localNGOs and community-based organizations orga-nized a program called Jagar in each of the 52relocation villages. The troupe prepared a packageconsisting of a “Prabhat Pheri” exhibition, songs,group discussions, and popular theater. Each of

these contained messages related to the earthquakerehabilitation program and invited people to sharetheir views, suggestions, doubts, or problems withthe team (TISS, 1997), but it did not include anytraining or beneficiary education in earthquake-resistant construction, since all the construction wasmanaged by contractors. This is in contrast to thereconstruction, repair, and strengthening program,which was an owner-managed program that devel-oped a number of strategies to educate villagersabout earthquake-resistant construction technology.

In retrospect, this lack of community education in therelocation villages was probably a weakness of theprogram. As villagers in the relocation villagesconstruct additions to their homes or new villagersadd homes, it is less likely that the owner-builderswill be familiar with earthquake-resistant technology.

In the CSSS study referred to above, only 15 percentof the relocation beneficiaries reported participatingin the community participation component of therebuilding program. The major areas of participa-tion for those who did participate included decidingthe location of villages, planning the layout ofvillages, and planning the allocation of houses(Figure 11).

Figure 11Interior of a newhouse in Killarivillage in theLatur district.


Building Technologies Usedin the New VillagesIn developing the technologies for the relocationvillages, the GOM’s prime concern was “to ensurethat housing provided in the area be earthquake-resistant to prevent any life loss due to possiblefuture earthquakes” (GOM, 1993). The followingfactors were considered:

1. Use of locally available materials (e.g., stone andsand) as much as possible.

2. Provision of permanent shelter to the affectedcommunities.

3. Functionality of the new houses and easymaintenance in the long run.

4. Thermal comfort.

5. Possibilities for replication of the proposedtechnologies in future construction.

Load-bearing masonry construction was consideredthe most feasible technology for the resettlement

villages in light of traditional construction, the levelof artisan skills, and various socioeconomic factors.The main objective of the housing design was tointroduce improvements in traditional constructionpractices, mainly by incorporating the provisions ofIndian seismic standards (such as IS 1893-1984,IS 4326-1993, IS 13827-1993, and IS 13828-1993)related to Seismic Zone IV of India. The key seismicprovisions used in the project were seismic concretebands (bond beams) introduced at the plinth, lintel,and/or roof levels (as per IS 4326-1993). In somevillages (e.g., Killari), vertical reinforcement barswere provided at the wall corners in the pockets ofthe hollow concrete blocks (Figure 12).

Certain improvements and innovations were alsointroduced in construction materials. Several typesof masonry products were offered for wall construc-tion, such as hollow/solid concrete blocks, stone-crete blocks (concrete blocks cast with larger piecesof stone rubble), and burnt clay bricks. Ultimately,the majority of houses in the relocation villageswere constructed using solid concrete blocks

Figure 12 Vertical reinforcement bars were provided in concrete block wallconstruction (Killari village, Latur district).


Figure 13 Construction of building foundations in Killari village. Stone masonrywas used up to the plinth level.

(200 mm [8 in.] thick) that appeared to be accept-able to the affected communities. Due to thevillagers’ immense fear of stone construction in theaftermath of the earthquake, stone boulders wereused only for the construction of strip foundations(up to the plinth level) (Figure 13).

Crushed stone rubble from the abandoned villagesites was also used for the production of concreteblocks. Also, because of the time constraint for theproject, mass use of stone masonry did not appearto be feasible, as building earthquake-resistant stonemasonry construction for 27,000 houses wouldcompel the shaping of hundreds of thousands ofstone boulders. On the other hand, the use of brickmasonry was discarded due to the limited availabil-ity of bricks (brick kilns) in the area; the bricksneeded to be transported from faraway locations. Infact, use of concrete blocks appeared to be the mostfeasible technology from the project managementviewpoint, as these units could be manufactureddirectly at the construction sites.

Reinforced concrete slabs were selected as feasibleroofing structures, replacing the traditional timberplank roofs. Mud mortar, frequently used as a binder

in masonry construction prior to the 1993 earthquake,was replaced with a cement-based mortar (1:6cement/sand ratio), as recommended by the GOMtechnical guidelines (1994b, 1994c), and by the GOIreports (GOI, 1993a, 1993b). It has been reportedthat, in an area with extremely hot weather and acutewater scarcity such as Marathwada, extensive use ofconcrete technology in the rebuilding presentedproblems related to construction quality (see LASA,1998; see also the section “Evaluation of ConstructionQuality and Program Outputs” in this report). De-tailed coverage of construction technologies used inthe relocation villages, including the house andvillage plans, is provided in the monographMaharashtra Emergency Earthquake RehabilitationProgramme–Housing Project (GOM, 1998b).

The CSSS study found that the vast majority ofrelocation village households responding (80 percent,or 14,543 households) believed the new houses to besafer. Of these, 88 percent said it was because of theuse of cement and sand, and 64 percent said it wasbecause of the use of earthquake-resistant technology(multiple answers were allowed). Of the 20 percent(3,472) of households that did not believe the house


to be safer, 79 percent said it was because of the fearof additional earthquakes and 70 percent had doubtsabout the quality of construction. It is interesting tonote, however, that even though the vast majority ofhouseholds believe the houses to be safer, most weresleeping outdoors even as recently as 1997, in partbecause of their fear of future earthquakes.

Special Case of Hybrid(Category B) VillagesObservation■ There were 22 villages in the GOM’s Category B

villages, which were not quite as severely dam-aged as the relocation villages discussed above,but which required complete rebuilding. Thesevillages were ultimately rebuilt in-situ, or in somecases, relocated nearby but with more benefi-ciary participation in the construction process.This modification in strategy appears to haveincreased satisfaction among the beneficiarieswith their new housing and resulted in actualincrease in floor area. Beneficiaries were able toconstruct larger houses by managing the con-struction process themselves.

DiscussionSixteen villages that were originally scheduled forrelocation because of severe damage were eventu-ally rebuilt in-situ. In these villages over 70 percentof the houses classified under Categories 4 and 5per the IAEE damage categorization (IAEE, 1986)suffered extensive damage. In subsequent consulta-tions with World Bank officials, and because ofsome of the negative consequences of movingvillages, it was decided to carry out the in-siturehabilitation of these villages.

The GOM decided to manage the rehabilitation ofCategory B villages through the Maharashtra Hous-ing Development Authority (MHADA), which was incharge of obtaining the consent for house recon-struction from the beneficiaries, preparing housedesigns, and managing the construction carried outby the contractors. As the plan was to carry out therebuilding in-situ, debris had to be removed prior toconstruction. However, the beneficiaries in thesevillages were not willing to give consent for the in-situ rebuilding; instead, they desired to move to

new sites like the beneficiaries of the relocationvillages. In October 1995, 17 months after theMEERP was launched, virtually no progress hadbeen made in the rehabilitation of these Category Bvillages (Nikolic -Brzev, 1995b). Due to the exten-sive damage in these villages and the fear of pos-sible earthquake tremors, villagers stayed in tempo-rary shelters (huts) located away from the villages.The temporary shelters, put up by the beneficiariesthemselves, had no water or power. The villagesappeared to be practically uninhabitable because ofa lack of maintenance. In spite of the difficult livingconditions, the beneficiaries were persistent in theirdecision not to have their houses rebuilt in-situ.

Based on interviews with the villagers, MHADA fieldstaff, and the PMU officials, the key reasons for theprolonged standstill in the Category B villages were:

• Changes in the rehabilitation policy. Initially theGOM planned to relocate these villages and thebeneficiaries were aware of that decision.

• Proximity to the relocation villages so it waseasy to see the new houses and amenitiesconstructed to better standards.

• Post-earthquake rebuilding was perceived as anopportunity to move away from a village. Thetraditional cluster architecture of the villagesmade it very difficult to carry out horizontalextensions of the existing houses. As a result,even before the earthquake many peopleconstructed their houses away from the villageson their own plots.

• In the majority of the Category B villages therewere at least two distinct groups: one groupdesired to stay in the old village and have theirhouses rehabilitated in-situ, while the othergroup desired to move out of the village.

• A considerable amount of debris in the villages.

In 1996, two years after the MEERP was launched,the GOM ultimately decided to have the Category Bvillages resettled at the new locations. However, itwas the responsibility of the beneficiaries to acquirethe plots for the new construction and only somecivic amenities were to be provided by the GOM—thus the “hybrid approach.” (The GOM agreed toprovide schools, electricity, and water supply, but notroads and drainage systems.) In three or four villages


were constructed in stone masonry. Cement-basedmortar was used in the construction of the founda-tions and the superstructure.

The beneficiaries themselves constructed theremaining houses in the village (Figures 14 through17). Each house was of a unique design, created bythe beneficiary and his family. In most cases, thewalls were constructed using brick masonry or, lessoften, stone masonry. Either concrete slab orcorrugated galvanized iron (CGI) sheets were usedin the roof construction. According to the rapidsurvey made by the authors, a considerably largerbuilt-in house area was obtained in the case ofowner-managed construction. In some cases, thebeneficiaries were able to construct four rooms withapproximately 450 sq. ft. of carpet area using thesame financial package. In this example, commu-nity-managed construction appears to have beenmore cost-effective (by approximately 33 percent)than mass construction.

Reconstruction, Repair, andStrengthening Program (RRSP)Observation■ The Reconstruction, Repair and Strengthening

Program (RRSP) was one of the most innovativeaspects of this project for two reasons. First,owners were directly involved in the constructionprocess. They were managing the process anddeciding whether to repair and strengthen orrebuild their houses and selecting which masonrymaterials were going to be used in construction(e.g., stone, clay bricks, or concrete blocks). Insome cases, beneficiaries and their families alsoprovided the labor or assistance in post-construc-tion activities (e.g., curing of concrete andmasonry construction). This resulted in a highlevel of satisfaction in these villages. Second, theproject incorporated knowledge about earth-quake-resistant technology and developed acomprehensive program to educate owners,artisans, and engineers through training, hands-on experience, demonstration projects, andmodel buildings.

with strong leadership, the owners took responsibilityfor managing their own construction. In other villagesthe beneficiaries came to the GOM and asked themto find NGOs for them. The GOM invited NGOs tomanage the construction in these hybrid villages. Intotal, 14 NGOs signed a Memorandum of Under-standing with the GOM. They were responsible formanaging the construction of 8,916 houses, and thefull financial assistance package was provided by theGOM. Construction of the remaining 1,712 houseswas managed by the owners, with assistance fromthe PMU, following the same guidelines and proce-dures as in the villages rehabilitated in-situ (GOM,1998c). There were also villages where the NGOsmanaged half the rebuilding and the owners man-aged the other half. During the last year of theproject another six villages were added to thiscategory, bringing the total of hybrid villages (relo-cated but managed by beneficiaries and NGOs) to 22.

These villages are examples of a hybrid housingrebuilding effort, both in terms of the managementstrategy and the construction standards. The villageswere rebuilt as a result of the mixed community-ledeffort and mass construction. In some cases theNGOs managed the construction carried out by thecontractors, with all the house plans predefined andthe same building materials used in all the houses.In other cases the beneficiaries managed the con-struction of the houses themselves, with the finan-cial and technical assistance provided by the gov-ernment in exactly the same fashion as in thevillages rehabilitated in-situ.

During a July 1998 field trip, the authors visitedseveral of these hybrid villages (Krimgold andNikolic -Brzev, 1998). The case of Jawli village in theLatur district appeared to be a particularly interest-ing one (Figures 16 through 19). Out of the 561 (100percent) houses in total, 215 (38 percent) houseswere constructed by an NGO (CARITAS), whereasthe remaining 356 houses (62 percent) were con-structed by the beneficiaries themselves (GOM,1998c). The entire village was resettled at a newlocation approximately 3 km away from the originalone. All the houses constructed by the NGO lookedalike; a single-house plan with two rooms andapproximately 250 sq. ft. of carpet area was used forthe entire village. Solid concrete blocks were usedin the wall construction, and a concrete slab wasused in the roof construction. Strip foundations


Figure 16 A beneficiary standing in front of hisold house in the Jawli village (Latur district).

Figure 15 (Left) Houses in the new Jawlivillage constructed by Caritas.

Figure 17 The same beneficiary standing in frontof his new house, constructed by Caritas (Jawlivillage, Latur district).

Figure 14 A house in the newJawli village constructed by thebeneficiaries. A larger built-inarea was obtained within thesame financial package for thisowner-managed project. Hutssimilar to that shown in theforeground were made bybeneficiaries who were afraid tostay in their houses, even thoughmany houses were onlymoderately damaged.


packages of financial grant assistance were offered forthese houses, depending on the level of damage. Dueto a number of defaulters (beneficiaries who hadinitially been included in the program and failed tocomply with the government policy and complete theconstruction per the GOM specifications) the totalnumber of beneficiaries in the program was approxi-mately 185,000.

An inherent complexity of the RRSP was the factthat it was a government-sponsored but community-managed program. The program was developed asa community-based construction process, controlledby the owners. A number of important decisionsrelated to the RRSP implementation were made bythe beneficiaries, their families, and entire villagecommunities.

Since damage to these houses was generally repair-able, the GOM’s initial approach was that the housescould have been rehabilitated or strengthened inorder to mitigate the losses due to future earth-quakes. However, the local population had a pro-nounced fear of stone masonry and even a year afterthe earthquake perceived stone as a culprit in thecollapse of or damage to their homes. Consequently,the GOM decided to offer an alternative for seismicstrengthening of the existing houses by grantingfinancial assistance to the beneficiaries for partiallyreconstructing their houses on the existing plinth(e.g., building a new room with earthquake resistanttechnology). The identical package of financialassistance was provided in both cases. The selectionof a preferred technology package was a volun-tary decision made by the beneficiaries.

Proposed Technology PackagesBuilding technologies proposed for the reconstruc-tion, repair, and strengthening of buildings in theRRSP were developed based on the lessons learnedfrom the earthquake, as well as from previous post-earthquake reconnaissance studies in developingcountries with similar construction practices.Construction techniques (not just the buildingmaterials) present one of the key factors determin-ing the performance of structures during an earth-quake. In the area affected by the September 30,1993 earthquake, there were many traditional stoneor brick masonry buildings constructed to higherstandards. While those buildings experienced

DiscussionThe RRSP sought to reconstruct, repair, andstrengthen approximately 212,000 moderately dam-aged houses scattered over 2,400 villages, in 13districts and covering 40,000 sq. km (15,440 sq. mi.)using earthquake-resistant technology appropriate forthe unreinforced masonry construction of the area.Ultimately, 189,000 houses were completed under thisprogram, since a number of beneficiaries failed tocomplete the construction according to the govern-ment specifications. This undertaking was one of themost complex and challenging components of thismassive rebuilding project. The beneficiaries took theinitiative to repair, strengthen, and reconstruct thedamaged houses with financial grants and technicalsupport from the government.

There were three categories of financial assistanceprovided to the beneficiaries in the RRSP. Ownerswere given the choice: if their homes were severelydamaged, they could completely reconstruct themusing earthquake-resistant technology; if moderatelydamaged, they could repair and strengthen theirhome, or they could build one new room attached tothe existing structure, using earthquake-resistanttechnology for the new room. The GOM financialassistance was limited to a fixed amount, given in theform of cash and kind. Using materials from olddamaged houses wherever feasible, the beneficiariessupervised construction work and, in some cases,provided family labor.

One of the major challenges of the rebuildingprogram was to introduce basic earthquake-resistantconstruction technology and know-how intononengineered rural construction practices. Toachieve this, the GOM provided hands-on training toall those involved in the RRSP implementation,particularly the beneficiaries, local artisans (especiallymasons), and engineers who were providing techni-cal assistance to the beneficiaries.

According to Nikolic -Brzev and Anicic (1994), it wasproposed that the houses that suffered varying degreesof damage in Maharashtra be either reconstructed orrepaired and rehabilitated in-situ. Out of those, themajority of houses (approximately 85 percent) wereslightly to moderately damaged (IAEE Damage Catego-ries 1 to 3), while the remaining 15 percent wereseverely damaged or had collapsed in the earthquake(IAEE Damage Categories 4 and 5). Two different


damage to the structural and nonstructural elements,they did not collapse and thus fatalities wereavoided. The lessons learned in this earthquakeindicate that the required level of seismic safety canbe achieved by using local construction materialsand skills in an adequate manner and by followingthe minimum seismic requirements recommendedby Indian seismic codes and standards (Figure 18).

Viable technology packages were offered within thetwo overall alternatives of the RRSP: 1.) Reconstruc-tion of the existing houses or portions thereof, and2.) Repair and strengthening of the existing dam-aged vulnerable houses. The Guidelines for Repair,Strengthening and Reconstruction of Houses Dam-aged in the September 30, 1993 Earthquake inMaharashtra, India (GOM, 1994b) and the Manualon Earthquake-Resistant Construction and SeismicStrengthening of Non-Engineered Buildings in RuralAreas of Maharashtra (GOM, 1998a) providedrecommended technologies for the RRSP implemen-tation. A summary of the salient features of therecommendations is presented below.

Reconstruction. In general, construction technolo-gies recommended for the in-situ rebuilding weresimilar to those adopted in the relocation villages.The following technologies were recommended forthe wall construction: UCR stone masonry, burnt

clay brick masonry, solid/hollow concrete blocks,stone-crete blocks (i.e., concrete blocks cast withlarge pieces of stone rubble), and stabilized soilblocks. Cement mortar was used as a binder inmasonry construction. The two recommendedroofing solutions were: corrugated galvanized iron(CGI) sheets, or reinforced concrete slab. Therecommended building technologies were alreadyin use in the area before the 1993 earthquake (to alimited extent in the villages and mainly in thetownships, e.g., district centers). The key featuresof the proposed rebuilding technologies were:

• Reinforced concrete bands (bond beams) at theplinth, lintel, and/or roof levels were recom-mended as one of the most important seismicprovisions per the requirements of pertinentIndian standards related to seismic-resistantdesign of low strength masonry (IS 4326-1993,IS 13828-1993).

• Replacement of mud mortar (which wasfrequently used in construction before theearthquake) with a lean (1:6) cement/sandmortar was recommended to ensure improvedseismic performance of new houses. Cementmortar also improves the health standard ofrural construction by providing protection fromwater penetration during monsoons (heavy

Figure 18 Anunreinforcedmasonry buildingconstructed to thehigher standardsand with a concretelintel band survivedthe 1993 earthquakewithout any damage(Killari village, Laturdistrict).


rains); hence the chance of dampness in theinterior of the house is minimized, a conditionwhich reportedly was a serious health hazard ofstone and mud construction.

• Construction using the same foundations andplinth, starting with the plinth band constructionwas recommended (Figure 19).

• In case of UCR stone masonry construction,recycling of the stones dismantled from theexisting house was recommended. Beneficiarieswere given guidance regarding the selection andcleaning of adequately shaped stones (Figure 20).

• Maximum utilization of the existing constructionelements, such as door and window frames, wasrecommended.

• Mandatory use of UCR stone masonry in animproved manner, with wall thickness restrictedto 450 mm (1 ft. 6 in.), through-stones, cornerstones, and shaped stone boulders. This was ofutmost importance to ensure considerableimprovement of the seismic safety of stonebuildings in future earthquakes.

Figure 20 A stonemasonry housereconstructed in animproved manner. Theround stone bouldersdismantled from theoriginal house were notused for the construction.

Figure 19 The GOM recommendedreconstruction on the same plinth. The existing

timber structure was preserved, and the wallswere reconstructed in the improved manner.


Repair and Strengthening. A considerable numberof houses in the RRSP suffered only moderatedamage. Hence, in the rehabilitation policy the GOMoffered a package of financial and technical assis-tance to beneficiaries whose houses were consideredfeasible for repair and strengthening. Apart fromoffering solutions for repair of earthquake-induceddamage, the package also included technologies tomitigate seismic hazards posed by the existingvulnerable masonry structures. Since the Marathwadaregion was previously not considered earthquake-prone, a large majority of houses were constructedwithout any seismic provisions and were vulnerable,as demonstrated in the 1993 earthquake. The mostimportant elements of the rehabilitation technologiesused in the RRSP were:

• The existing heavy roofs were either removedand replaced with lighter structures or, alterna-tively, roof weight was considerably reduced byremoving some of the heavy mud.

• Reinforced concrete bands (ring beams) wereinstalled at the lintel and/or roof levels in orderto preserve the integrity of the building in anearthquake and to minimize chances of an out-of-plane wall failure (Figure 21).

• Stone masonry walls were strengthened byproviding through-wall anchors (through-stones)at certain predefined locations.

• Knee-braces were provided at the beam-to-postjunctions to prevent the swaying of timberframes (Figure 22).

Beneficiary-Managed Selection ofConstruction TechnologiesThe RRSP was designed to be managed by thehomeowners themselves. They were expected todecide whether to strengthen or to construct a new“safe” room in their house. They were encouraged to

Figure 21 An example of concrete band installation at the eaves level of a stone masonry wall.


Figure 22 A beneficiary of a rebuilt strengthenedhouse where a timber frame was retrofitted with kneebraces (steel angles).

participate in the construction themselves, and theyreceived training in the principles of earthquakeresistant construction from the GOM Junior Engineers(JEs), community participation consultants, and someof the local village leaders (Figure 23).

More than one million people (including the benefi-ciaries and their families) participated in the RRSP.While the large majority of the beneficiaries were of asimilar rural background, there was a broad range ofeconomic and social statuses. The two most impor-tant decisions made by the beneficiaries/owners ofmoderately damaged houses (IAEE damage catego-ries 1 to 3) in the RRSP were to choose a constructiontechnology package (i.e., reconstruction or strength-ening), and to choose the wall/roofing material to beused in the reconstruction. The rationales for thebeneficiaries’ choices and preferences are discussedin the following sections.

Selection of MasonryTechnologies in RebuildingIn the initial phase of rehabilitation, the largemajority of beneficiaries refused to use stone in theconstruction. Instead, they showed a strong prefer-ence for alternative, “modern” masonry technolo-gies, like burnt clay bricks or solid concrete blocks.According to a field survey (Nikolic -Brzev, 1995),

Figure 23 A largemajority ofbeneficiariesdecided todismantle theirstone houses andcarry out newconstruction inbrick masonry.


the majority of beneficiaries in the RRSP villageswho chose the reconstruction option selected burntclay bricks for the wall construction instead ofstone, although in most cases they had owned stonemasonry houses before the earthquake. The survey(based on approximately 200 villages) revealed thatover 75 percent of the beneficiaries in the Laturdistrict and 60 percent in the Osmanabad districtchose clay bricks for the wall construction (Figure23); the second most popular masonry material wasstone (Figure 24). Other types of masonry units(such as solid concrete blocks, stone-crete blocks,or soil cement blocks) appeared to be less popular.

The beneficiary preference appears to be mainly ofa psychological nature. The beneficiaries thoughtthat in case of a wall collapse during an earthquake,the chances of a serious injury (or even fatality) in astone masonry house would be greater than in asimilar house made of clay brick or concrete blockwall construction. Interestingly, the beneficiariesbelieved that stone boulders were more hazardousthan clay bricks or concrete blocks because stoneboulders are larger than the other two types ofmasonry units. This belief was widespread amongthe beneficiaries (Nikolic -Brzev, 1996; Krimgold andNikolic -Brzev, 1998).

In fact, however, the progressive collapse of amasonry wall usually happens suddenly because

masonry has a brittle nature. Since the weight ofmasonry rubble from a collapsed wall would besimilar no matter what masonry materials were used(it would mainly be a function of the wall thick-ness), the complete collapse of any masonry wallwould likely result in a similar kind of casualty. Inmany cases, the beneficiaries decided to use stonemasonry up to the sill level, and then burnt claybrick masonry for the upper wall portion. In mostcases, however, stones were used for the construc-tion of strip foundations up to the plinth level.

According to the GOM policy for the RRSP villages,the selection of building construction materials wasto be made at the discretion of the beneficiariesthemselves (from technology options offered in theGOM Guidelines – GOM, 1994a, 1994b). It wasinitially anticipated that a majority of the benefici-aries would select stone masonry as a preferredchoice. As suggested in the GOM Guidelines (GOM,1994b), the beneficiaries could recycle the stoneboulders from their original houses. On the otherhand, in order to carry out brick masonry construc-tion, the beneficiaries had to procure bricks at themarket rate, which doubled in the four years ofproject implementation.

There were also cost considerations to rebuildingusing stone masonry construction and upgrading tothe level specified in Indian seismic standards. The

Figure 24 In somevillages, stone masonryremained the mostpopular building materialafter the earthquake. Herea Junior Engineer isinspecting several newstone masonry houses.


stones had to be cut and shaped, andthe stone masonry used a highervolume of cement because of the largerspaces between stones (Figure 25). As aresult, stone masonry appeared to bethe most expensive constructiontechnology of all the packages offeredin the MEERP. This cost, therefore, wasanother important factor that led themajority of beneficiaries to select brickmasonry construction, even though italso had increased in cost.

Finally, it appears that social factorsalso played an important role in theselection of masonry material. Themajority of the population in theearthquake-affected Marathwada regionwas poor. A brick masonry house wassomething most had aspired to but hadnever been able to afford. Brick ma-sonry construction in cement mortar islocally known as “pukka” constructionand is found mainly in the townships(such as in the Latur or Osmanabaddistrict centers). Prior to the earth-quake, pukka houses were generallyowned by the wealthier members of thevillage (landowners or tradesmen).Many beneficiaries used this program asa unique opportunity to develop andconstruct “dream” houses (Figure 26).As the financial grants offered by theGOM were of a set amount, manybeneficiaries contributed their ownfunds in order to expand the size oftheir new houses. The typical benefi-ciary contribution was 10 to 20 percentof the GOM grant assistance.

Figure 25 New stone walls were constructed usingshaped stones cut by stonecutter artisans.

Figure 26 New houses were constructed to much higher standardsthan the houses that were built before the earthquake. An old house(left) and a new house (right)are shown here, with the beneficiarystanding in front of the new one.


Strengthening of ModeratelyDamaged Houses vs. NewConstruction of a “Safe” RoomFrom the inception of the RRSP, it was apparent thatstrengthening was not a preferred technologypackage for the beneficiaries. Only a very limitednumber voluntarily selected strengthening over newconstruction of an additional room. The officialfigure is not known; however, according to theQuality Assurance and Technical Audit consultants(LASA, 1998), only 0.1 percent of the beneficiariesdecided to repair and strengthen their houses. Basedon numerous field visits and interviews with hun-dreds of beneficiaries, field engineers, PMU officials,and community participation consultants held from1994 to 1998 (Nikolic -Brzev, 1999; Krimgold andNikolic -Brzev, 1998), the authors believe that themost important factors influencing a beneficiary’schoice of construction technology package (con-

struction of a new room versus strengthening of theentire existing house) were:

• House condition and/or age. At the time ofthe earthquake, many houses in the affectedvillages were in a deteriorated condition due toaging, adverse weather conditions, and lack ofmaintenance. In many cases, for example, forhouses constructed in sun-dried mud/adobeblocks, strengthening was not considered afeasible rehabilitation option (Figure 27).

• Level of beneficiary income. A large majorityof low-income beneficiaries selected reconstruc-tion as the preferred option. Before the 1993earthquake, a large majority of low-incomebeneficiaries owned hut-like (“kutcha”) houses,for which strengthening was not a feasiblerehabilitation option in any case. On the otherhand, middle- and upper-class beneficiariesowned houses that were only 15 to 20 years old

Figure 27 An example of a situation where retrofitting was not a feasible option dueto excessive wall thickness and poor quality of construction. Masons are shownconstructing a new plinth band.


and were better constructed. They were builtwith cut stones instead of round field stoneboulders, lime or cement mortar instead of claymud mortar, and quality timber in the frameconstruction. For these reasons, upper-incomebeneficiaries owned a majority of the housesstrengthened in the RRSP.

• Need for an expansion of built-in housearea. The area affected by the earthquakeconsisted of old rural settlements inhabited forseveral centuries. Many houses in the villageswere constructed hundreds of years ago(Nikolic -Brzev, 1999). At the time of theiroriginal construction, those dwellings had beendesigned to shelter smaller families. Due to thespecific architectural and structural features oftraditional houses and the limited plot areaavailable in the congested village environment, itwas virtually impossible to make either horizon-tal or vertical building extensions. The need toextend the existing built-in house area wascertainly one of the underlying reasons for manybeneficiaries to select reconstruction instead ofstrengthening as an option. This was particularlycommon for beneficiaries whose houses hadonly been moderately damaged (Figure 16).

• Extent of the earthquake damage/proximityto the epicenter. For the beneficiaries whosehouses suffered more extensive damage in theearthquake (e.g., collapse of a portion of a stonemasonry wall or large gaping cracks in a wall),strengthening was an unacceptable option. Dueto factors such as relative proximity to therelocation villages where mass construction ofbrand new houses was in progress and memo-ries of the earthquake that had caused a largenumber of fatalities because of the poor qualityof stone masonry construction, it was verydifficult for the GOM field staff to persuade thebeneficiaries to re-inhabit their houses. In theyear following the earthquake, fear of a recur-rence was so strong that the majority of benefi-ciaries, especially the ones whose villages werelocated close to the epicenter, stayed in self-made bamboo and thatch huts at night (Figure14), even though they occupied their housesduring the day. In spite of all the GOM educa-tion and information dissemination campaignsand the special incentive schemes developed to

promote the strengthening option, the majorityof beneficiaries in the villages located closest tothe epicenter ultimately decided to reconstructrather than to strengthen their houses.

• Time constraint for the recovery. Accordingto the conditions of the World Bank credit,MEERP was developed as a 3-year emergencyprogram. Therefore, the GOM officers who weretaking part in the RRSP were pressed to meetdeadlines to complete housing construction ontime. Moreover, a majority of the JEs who wereproviding technical guidance and overseeing theprogram implementation at the “grassroots”village level had been hired on a contract basis,with a 6-month contract extension period. Theywere given a certain target number of houses torehabilitate in 6 months, or else their contractwould be terminated. Besides the penalty clause,the JEs were also given financial incentives ifthey completed more than the target number ofhouses. Under such circumstances, it was mucheasier and more time-effective for them tooversee reconstruction rather than the strength-ening of a house. Strengthening is clearly a morecomplex option, both in terms of technicalsolutions (each house was a unique case study)and physical implementation. A considerablylarger effort was required by a JE to providefrequent site instructions to the artisans whowere generally not familiar with the strengthen-ing technology.

• Implementation experience. The RRSP wasofficially launched in June 1994. The PMUTechnical Guidelines (GOM, 1994b, 1994c) werepublished that month, and subsequently severalorientation training sessions for the PMU engi-neering staff took place. In September 1994,over 400 JEs were mobilized and started fieldactivities in the villages. The RRSP strategydocument (Nikolic -Brzev and Anicic, 1994)proposed phasing the program into four stagesand the construction target numbers weresubsequently modified in the MEERP Implemen-tation Plan (GOM, 1995b). Per the initial agree-ment of the World Bank credit (World Bank,1994), the RRSP (and MEERP in general) wasscheduled for completion in June 1997. How-ever, due to various problems, the programimplementation progressed slowly in the first


from the higher economic status whose houses hadnot been severely damaged (Nikolic´-Brzev, 1999).

Figure 28 shows that the pace of progress was veryslow in the initial two years of the program imple-mentation and the problems were numerous, in partbecause of the community-managed strategy for theprogram implementation. These problems included:

1. Lack of beneficiary motivation to participate.Initially the beneficiaries were also hoping to geta “better” package of financial assistance, i.e., toresettle at the new locations like the beneficia-ries in the relocation villages.

2. Beneficiaries’ lack of awareness of the deadlinefor the rehabilitation effort.

3. Administrative problems related to the disburse-ment of cash installments.

4. Beneficiaries’ reluctance to carry out constructionon the existing plinth (they wanted to rebuild ontheir own plots away from the village).

5. Increased foundation costs in the expansive soilareas (black cotton soil).

two years (from June 1994 to 1996) and thus itwas not possible for the GOM to complete thiscomponent within the initially agreed upon timeframe. Consequently, the World Bank agreed toextend the entire MEERP duration by 18 months,until December 1998.

Figure 28 summarizes the physical progress of theRRSP implementation (expressed in terms of thenumber of completed houses) in the last four years.It is very interesting to compare the actual pace ofprogress with the original implementation plandeveloped by the PMU. According to the field reports(LASA, 1998), it took beneficiaries between 3 and 24months (per house) to complete the post-earthquakerehabilitation sponsored by the GOM. At the end ofthe RSSP implementation, approximately 10 to 15percent of the total number of beneficiaries weredesignated as “defaulters.” The defaulters were thosewho had consented to carry out the constructionunder RRSP according to the GOM policy and hadreceived at least one (or more) installments of thefinancial assistance but had failed to complete theconstruction during the RRSP implementation period.Interestingly, the defaulters were mostly beneficiaries

Figure 28 Construction progress in the repair and strengthening villages.


6. Water scarcity (an acute problem).

7. A shortage of local construction labor availablefor the RRSP.

8. Management problems such as inadequatecoordination between the PMU and the engi-neering and administrative field staff.

The PMU and its consultants undertook a series ofactions to deal with the problems that hampered theprogress of the RRSP implementation. Theseincluded:

1. Development of the implementation work planat the micro (village) level (developed by thePMU and the project management consultants).The plan was displayed in each village with thetime frame and deadline for the beneficiaries ina particular village.

2. Development of an improved strategy formonitoring progress, including use of thecomputerized PMIS by the project managementconsultants.

3. Decision by the GOM to allow new constructionon the land acquired by the beneficiariesthemselves and away from their original houses.

4. Improved monitoring strategy for material dis-bursement by distributing material coupons withan expiration date (45 days from the release).

5. Improved strategy for community participationand information dissemination, including theappointment and training by the communityparticipation consultants of community-basedorganizations (Mahila Mandals) who wereresponsible for information dissemination at thevillage level.

6. Improved management strategy for the JEs ap-pointed on a contract basis, including a perfor-mance-oriented incentive/penalty scheme.

7. Improved accounting procedures and recruitmentof additional accounting staff at the field level.

8. Improved troubleshooting procedures, whichincluded formation of special PMU teams toappraise villages where implementation prob-lems were reported.

Training InitiativesA number of important innovations were necessaryto successfully implement this complex rebuildingprogram, including various types of training andeducation initiatives. This section discusses thetraining initiatives developed as part of this project,with the major observations connected to eachinitiative highlighted in italics at the beginning ofthe discussion.

Engineers as Technical Advisorsto the BeneficiariesObservation■ The technical support and guidance provided to

the beneficiaries and local artisans by the PMUengineering staff were of great importance toensure the successful transfer of building technol-ogy know-how to rural residents who had noprevious background in earthquake-resistanttechnology.

DiscussionThe RRSP program used Junior Engineers (JEs),primarily from the earthquake-affected area, to workwith individual beneficiaries to provide technicalassistance in the reconstruction or repair and strength-ening of their homes (Figure 29). These engineerswere each assigned to one or more villages and wereexpected to help beneficiaries work out the design forrepair and strengthening or for new construction. Theyalso acted as intermediaries between the GOM and thelocal villages. Activities of the JEs in the Latur andOsmanabad districts were supervised by the PMUsenior engineering staff, including deputy engineers,executive engineers, and superintending engineers,who visited the villages regularly and communicatedwith the beneficiaries and village leaders.

As mentioned earlier, in the RRSP the engineerswere given the unique role of technical advisors tothe beneficiaries, who then actually managed therebuilding of their own houses. JEs were hired bythe GOM on a contract basis to ensure an adequatelevel of technical support and supervision in thevillages where rebuilding in-situ was in progress. JEseither resided in the villages allotted to them orcommuted daily by bus, motorcycle, or bicycle. Themain scope of their work was to:


As the work began in each village, JEsorganized a village gathering to explain theproject and the technology packages offeredto the beneficiaries. Later, JEs visited eachbeneficiary and discussed the programaccording to the predefined implementationprocedures (GOM, 1994e, Nikolic -Brzev andAnicic, 1994).

Since the traditional masons did not have anybackground in earthquake-resistant construc-tion technology and since some of them hadvery limited construction skills in general, JEsalso had to hold training sessions for the localmasons. The JEs demonstrated excellent socialskills, partly because their backgrounds weresimilar to the beneficiaries.

There were 60 female JEs out of a total of790, considered a positive feature of theprogram, given the feudal nature of the area.The female JEs had an advantage over themales in being able to communicate with thefemale beneficiaries (because of specificsocial norms in the rural areas of India).Hence, progress in the villages supervised bythe female JEs was either equal to or betterthan the pace in the majority of villagessupervised by male JEs. Due to a specialincentive scheme developed by the PMU, JEswere financially motivated to accomplishmore work than their target quota within asix-month contract period. It is noteworthythat six-month renewable contracts for JEswere used as a strategy to avoid their de-mands for permanent positions with the GOM(at the time of the earthquake, there was ahigh unemployment rate among civil engi-neers in Maharashtra). In spite of that, JEs

went on three one-month strikes during the pro-gram implementation, thereby causing disruptionand delays in the project. At the end of the project,the majority of PMU engineers were pleased that theyhad had an opportunity to provide a social serviceto their communities, especially since a considerablenumber of them were originally from the earth-quake-affected area (Nikolic -Brzev, 1999, Krimgoldand Nikolic -Brzev, 1998) (Figures 30 and 31).

1. Prepare construction cost estimates for eachhouse, including any estimates of additionalfunds required by a beneficiary.

2. Process the building materials entitlement andcertify their use.

3. Ensure that construction was carried out inconformance with the GOM Technical Guide-lines (GOM, 1994b, 1994c).

4. Oversee progress in the construction and certifythat a beneficiary was qualified for the nextinstallment of financial assistance.

Figure 29 A GOM Junior Engineer inspecting a newlybuilt stone wall and pointing out a through-stone.


Figure 30 JuniorEngineers inspecting aretrofitted stone house

that has through-stonesand a roof band.

Training of Junior EngineersObservation■ Through a comprehensive training program

developed as part of the RRSP component of therebuilding project, engineers working in the fieldwere given an opportunity to improve theirknowledge of earthquake-resistant technology forthis and future projects.

DiscussionThe key technicians in the RRSP implementationwere JEs hired by the GOM to oversee the projectfield implementation and provide technical guidanceto the affected village communities. Most of the JEswere recent graduates of the local engineeringcolleges with limited backgrounds in the area ofearthquake engineering. Due to a rather high rate ofunemployment among civil engineers, they werehappy to have the opportunity to work. Training ofthis engineering field staff was mainly a responsibil-ity of the national seismic consultants. The trainingcurriculum was primarily based on the materialcovered in the Guidelines for Repair, Strengtheningand Reconstruction of Houses Damaged in the

September 30, 1993 Earthquake in Maharashtra,India (GOM, 1994b, 1994c). The guidelines werethe first technical publication issued by the PMU inthe project implementation phase. A major part ofthe guidelines concentrated on the technologies andtechnical specifications for the repair and strength-ening of moderately damaged buildings that werewithin the scope of the RRSP. The document in-cluded typical damage patterns reported in the 1993earthquake (illustrated with photographs), theassociated seismic risk, and the methodology forrepair and retrofitting. Since approximately 80percent of the damaged houses were of stonemasonry construction and the remaining portionwere mainly unreinforced brick masonry, theguidelines concentrated on those two types. Alongwith the remediation measures, the constructiontechnology packages that were to be used in thereconstruction were also included.

The guidelines were issued in English and severalthousand copies were printed. They were distrib-uted to PMU engineering personnel and to thePublic Works Department engineering staff workingin other rural areas of Maharashtra. An abridged


version of this material was translated into the localMarathi language, printed in August 1994 anddistributed to most beneficiaries and village leaders.In order to facilitate field implementation of thetechnologies recommended in the guidelines, thePMU engineering staff prepared a booklet thatincluded unit cost estimates for the key strengthen-ing and reconstruction features (GOM, 1994f).

In addition to the guidelines, the training materialincluded information on basic concepts of earth-quake engineering and the design of timber struc-tures. The consultants used printed materials andvideo presentations in the training sessions. Onseveral occasions consultants prepared technicalnotes on mistakes observed in the constructioncarried out in the RRSP. Apart from the seismicengineering consultants, quality assurance andtechnical audit consultants also conducted severaltechnical workshops for the PMU field engineers.Curriculum for those workshops included topicsrelated to the quality of cement-based constructiontechnologies, as well as common mistakes reportedin the field implementation.

In the initial phase of the program implementation,the consultants organized short-term (one to twodays) training courses for the PMU engineering staff.The courses were conducted once a month in thedistricts of Latur and Osmanabad. Occasionally,training courses were also organized in other earth-quake-affected districts, especially Solapur and Satara.A special series of training courses was organized inthe 13 districts of Maharashtra that were not damagedin the 1993 earthquake, where implementation of thePilot Strengthening Program (PSP) was occurring (seediscussion on page 49).

Training of ArtisansObservation■ This program was structured to ensure the

improvement of existing construction by providingtraining to the technicians involved in programimplementation, including the construction laborand engineering staffs. The training program inthe MEERP was customized. This training initia-tive contributed to the success of the project,particularly the in-situ rebuilding component.

Figure 31 A JuniorEngineer (in blacktrousers) with abeneficiary standingin front of acompleted house inthe repair andstrengtheningprogram.


DiscussionIn order to ensure the successful transfer of knowl-edge regarding earthquake-resistant constructiontechnology to the rural communities, the GOMdeveloped training programs for artisans in theearthquake-affected area. Early in the projectpreparation phase (Nikolic -Brzev and Anicic, 1994),the GOM realized that the number of traditionalartisans (especially masons) working in the areabefore the earthquake (approximately 2 percent ofthe population) was not adequate to meet thedemands of the in-situ rehabilitation of 185,000damaged houses. In addition, the program was tobe managed by the local communities and the

construction carried out in large part by localartisans (Figure 32). Hence, the GOM launched thefollowing three training initiatives for masons:

• Training of unskilled labor. In June 1994 theDirectorate of Vocational Education and Traininglaunched training programs for unskilled labor inthe earthquake-affected area; the programs weresponsored by MEERP. An existing network of 32vocational training centers in the most affecteddistricts of Latur, Osmanabad, Satara, andSolapur was used for this purpose.

The training lasted two months and coveredfour trades: masonry, carpentry, electric works,and welding. The training curriculum covered

Figure 32 A village artisan preparing to install a through-stone. Here, he isremoving a stone with a crowbar.


masonry skills for basic elements of stone andbrick masonry construction, as well as elementsof earthquake-resistant masonry construction.Training for each trade was two weeks. Trainerswere the instructors of the Industrial TrainingInstitutes, a chain of government centers incharge of providing education to unskilledlaborers. Prior to the training, the trainersthemselves were given training in earthquake-resistant construction.

The multitrade training course was organizedbecause even though MEERP only requiredtraining for the masonry trade, interest inmasonry training was very poor. Of 3,500candidates interviewed, less than 10 percentshowed interest in being trained in masonryskills. Another problem was that the GOM wasnot in a position to bind the trainees to takepart in the RRSP because it was a community-managed effort, and it was up to the individualbeneficiaries to contract for the constructionlabor. Nevertheless, from May 1994 to May 1995over 6,800 individuals were trained under thisprogram (Nikolic´-Brzev and Anicic, 1994).

• Training of traditional masons in the Laturand Osmanabad districts. The PMU made aneffort to improve the skills of the traditionalmasons by imparting practical, hands-on trainingin earthquake-resistant construction technologyfor masonry buildings. The idea for this trainingemerged during the RRSP implementation. Thecommunity participation consultants proposedthe program in May 1995 and it started inNovember 1995. A team of one deputy engineerand two JEs was assigned to this training full-time; the training was managed by a PMUexecutive engineer (Deshmukh, 1998). Thetraining curriculum was developed by thecommunity participation consultants and re-viewed by the chief engineer, PMU, and theforeign seismic consultant.

The two-day training covered basic principles ofstone masonry construction, use of cement-based mortar, and the construction of keyseismic features for masonry buildings.

The first part of the training was conducted inthe classroom; the second part was hands-on

practical training conducted in the villageswhere construction under the RRSP was inprogress, so the trainees got an opportunity tocarry out actual construction during the training.During the classroom training, the trainees wereshown patterns of damage to masonry construc-tion reported in the 1993 earthquake; commondrawbacks in traditional masonry constructionfound in the earthquake-affected area werehighlighted during the training. Each traineereceived a mason tool kit and a stipend equal totwo days at market wages for a skilled mason.

At the end of the training, each trainee receiveda certificate and took an oath vowing that hewould follow the principles of earthquake-resistant construction that he had been taught.In the period from November 1995 to February1997, approximately 4,000 traditional masonswere trained in the Latur and Osmanabaddistricts under this program. Most of themparticipated in the RRSP implementation. Therewas an obvious improvement in the quality ofmasonry construction in the villages wheretrained masons were doing construction. Themasons were very proud of the training andliked to show their certificates to visitors(Nikolic -Brzev, 1999).

• Hands-on mason training in the villages bythe PMU engineering field staff. One of themost important training initiatives was undoubt-edly the training provided to traditional artisansby the PMU engineering staff working in thevillages. Most important, JEs of the PMU playeda very important role in educating the localartisans regarding the improved constructionpractices.

The JEs taught local artisans how to construct aseismic band, an important seismic feature veryrarely found in the area before the earthquake.Thanks to the JEs, local artisans learned to bendsteel reinforcement bars and manufacturekneebraces for the strengthening work (Figure33). In addition to the JEs, their peers in thePMU, deputy engineers, executive engineers,superintending engineers, and chief engineersalso visited the villages and provided guidanceto the traditional artisans on a regular basis.Figure 34 shows a PMU deputy engineer (with a


Figure 34 A PMU Deputy Engineer (left, wearing a scarf) explains theconcept of seismic band to traditional artisans (masons).

scarf) explaining theconcept of a seismic bandto a homeowner and themasons. “If four personsare braced together (like ahouse with a seismicband), they become moreresilient to the effect of apush than each of the fourpersons individually.”

Figure 33 A villageartisan working onfabrication of steelhoops for seismicbands and knee-bracings (rolled steelangles at the right).


Technical AssistanceObservation■ Since the implementing agency (GOM) did not

have any previous experience in managing apost-earthquake rehabilitation project on such ascale, technical assistance played an importantrole in project implementation. An importantinnovation that the GOM intends to use in othergovernment projects is the program managementconsultant.

DiscussionConsidering the complexity and scale of the project,at the time of negotiations with the GOM, the WorldBank required that a number of external consultantsand agencies be appointed in the MEERP preparationand implementation. The World Bank outlined anumber of consulting assignments consideredimportant for the design, supervision, and monitoringof project components (World Bank, 1994a):

1. Project management experts.

2. International earthquake engineering experts todesign the strategy for the reconstruction in-situand propose building technologies for recon-struction, repair, and strengthening of existingbuildings.

3. Local earthquake engineering experts to providethe PMU technicians with training related toearthquake-resistant construction technology.

4. Consultants experienced in earthquake engineer-ing to investigate and test, both in the laboratoryand in the field, stone masonry strengtheningsolutions to be adopted in the affected areas.

5. Architectural, engineering, and planning consult-ants to prepare detailed architectural andengineering designs and contract documentationfor the relocation villages.

6. Consultants for community participation.

7. Consultants for quality assurance and technicalaudit functions.

8. Consultants (national and foreign) for develop-ing the State of Maharashtra disaster manage-ment plan.

9. Experts to train village artisans, builders, andNGOs.

10. Experts to prepare educational video films forconfidence building and public awareness.

Except for two foreign seismic engineering consult-ants (sponsored by the World Bank credit), threedisaster management consultants (sponsored by theDFID), and a project management advisor (spon-sored by the ADB), all of the remaining consultingservices were provided by Indian consulting firms,government agencies, or individuals. Technicalassistance was also provided by consultants in thepreparation of rehabilitation action plans for thosewhose lands were acquired and thus becamelandless, the documentation of the entire projectexperience, and the development of databases ofbeneficiaries and disaster management information.

Of particular importance was the technical assis-tance provided by the program management con-sultants, who set up a management informationsystem, the monitoring of performance indicators,and a monthly and yearly reporting mechanism. Inaddition, valuable technical assistance was providedby the quality assurance team. This was the firsttime that the state used an external mechanism forproject management and quality supervision on aregular basis. The GOM hopes to incorporate suchassistance in future project management activities.

In total, approximately $21 million (10 percent ofthe total World Bank credit) was allocated fortechnical assistance (GOM, 1998b). This represents avery high percentage of the total project cost for atypical World Bank project and indicates the highlevel of importance placed on technical assistance inthis project. In addition, a number of technicalsupport and training activities in MEERP weresponsored by grants provided by the Departmentfor International Development (approximately U.S.$3.8 million), the United Nations DevelopmentProgram ($0.6 million), and the Asian DevelopmentBank ($0.6 million).


Innovative Strategies toDemonstrate Earthquake-Resistant ConstructionA major strength of this rebuilding project was thatit emphasized the introduction of earthquake-resistant technology and mitigation or reduction offuture losses through the strengthening of bothdamaged and undamaged structures. In order tointroduce the new technology and demonstrate theefficacy of mitigation, a number of innovativedemonstration strategies were adopted:

• Confidence-building project

• Model buildings

• Pilot demonstrations

• Base isolation demonstration project

Each of these strategies is discussed briefly below.

Confidence-Building ProjectObservation■ Since the population of the affected area lost

confidence in stone as a building material, theGOM supported an innovative project. Tests wereconducted in the field on a rudimentary shaketable to demonstrate to beneficiaries and techni-cal professionals the reliability of strengtheningtechnologies for stone masonry construction.

DiscussionDue to the poor performance of stone masonrybuildings in the 1993 earthquake, the population ofthe affected area lost confidence in stone as abuilding material. They considered it the primaryculprit responsible for the loss of life and destruc-tion of their houses. Similarly, in the initial phase ofthe post-earthquake rebuilding, they were generallysuspicious of the effectiveness of seismic strengthen-ing technologies proposed for the mitigation ofseismic hazards attributed to stone masonry build-ings. In spite of all their efforts, the GOM field stafffound it difficult to convince the beneficiaries thatthere was nothing wrong with stone as a buildingmaterial if it was used in a proper fashion. However,the fear was so pronounced that many people weresleeping in thatch huts outside their houses for

almost a year after the earthquake. Some of thehouses that were retrofitted in the initial phase ofMEERP were either completely deserted (used asstorage) or inhabited only in the daytime.

In order to build confidence among the localpopulation and the GOM technicians concernedabout the seismic safety of retrofitted stone masonrybuildings, the GOM launched the ExperimentalVerification and Confidence Building Project, carriedout under the leadership of Professor A.S. Arya ofthe Earthquake Engineering Department, Universityof Roorkee. The program was sponsored by theADB and the World Bank. The most important testsconducted under the project were impact shake-table tests on stone masonry building models (ADB,1995, Arya, 1996). Two one-room house modelswere constructed to a reduced scale (approximatelyhalf of the real size) using local building materialsand following the construction practices typical forthe Marathwada area. Defects in traditional construc-tion practices, such as using round stone boulders inthe walls, the absence of stone interlocking andthrough-stones, very thick walls, unbraced timberframes, and a thick mud overlay atop a roof werereplicated in both of the models. For testing pur-poses, however, one model was retrofitted using theseismic strengthening methodology recommendedin MEERP (GOM, 1994b, 1998a) and the other wastested in its unstrengthened condition (Figure 35).

The tests were carried out using a custom-mademechanical shaking table facility in the city ofOmerga, in the Osmanabad district. Both modelswere tested simultaneously using tractors thatcreated repeated shocks to the shaking table. Thespecimens were subjected to 12 shocks in total. Theintensity of shocks varied from 0.06g to 1.6g interms of acceleration level (g denotes the accelera-tion due to gravity). Major findings of the testingprogram are summarized below (ADB, 1995).

• The strengthened specimen sustained the effectsof all 12 shocks without collapse. The last shockwas extremely severe, characterized by peakbase (table) acceleration of 1.6g. The modeldeveloped only minor cracks in the shockcharacterized by a peak base acceleration of0.47g. The extent of cracking increased inproportion to the intensity of impact loading,although collapse did not occur during the tests.


Figure 35Experimentalverification of theretrofitting technologyat Umarga,Osmanabad district. Aretrofitted model isshown at the right andthe unstrengthenedmodel (damaged) isshown on the left.

• The unstrengthened specimen developed severecracks. Walls delaminated during the shock—characterized by a base acceleration of 0.47g;masonry fell off the walls and wooden frametilted in the subsequent shocks. The damagepattern observed in the tests resembled thatobserved in the 1993 earthquake.

It is very difficult to draw general conclusions onthe performance of an entire building categorybased on a single experiment, especially wherenonengineered masonry buildings are concerned.Masonry construction in general, but especiallystone masonry, is characterized by broad variationsin construction quality.

Consequently, the seismic performance of masonry isaffected by the mechanical characteristics of thismaterial (with significant implications as to itscapacity to sustain lateral loads) as well as how it isused in construction. In addition, it is very difficult toscale the properties of stone masonry because of theunique composition and geometry of each stone wall.Finally, impact-type shake table tests represent onlycrude simulations of real earthquake motions; thefrequency content and the duration of an earthquakeare equally important factors in building response to

the maximum intensity, expressed through the valueof peak base acceleration. It is difficult, if not impos-sible, to account for frequency and duration usingtractor-induced (impact) dynamic motions.

The confidence building program undoubtedlyserved its purpose—it demonstrated the effective-ness and the benefits of strengthening technologiesproposed in MEERP. The tests were carried outpublicly and were attended by more than 300 GOMengineers, district administrators, and beneficiariesof the MEERP. The tests were also recorded onvideo and were used by the community participa-tion consultants working in the RRSP villages.According to interviews with the PMU field engi-neers (Krimgold and Nikolic -Brzev, 1998; Momin,1998), the tests convinced the GOM technicians ofthe effectiveness of the proposed strengtheningtechnologies. Based on the information provided bythe PMU engineers, the beneficiaries who attendedthe tests were also convinced of the effectiveness ofstone masonry strengthening technologies. How-ever, very few beneficiaries (including those whoattended the tests) reversed their original decisionsto build a new room rather than strengthen theexisting house.


Model BuildingsObservation■ To demonstrate earthquake-resistant building

construction technology in the earthquake-affected area, the GOM built model buildingsthroughout the region as part of the rebuildingprogram. They were built within the first year ofthe project implementation and featured earth-quake-resistant technology and the changes inconstruction practices.

DiscussionIn the initial year of the post-earthquake housingrebuilding, over 500 model houses were constructedall over the affected area to demonstrate cost-effective building techniques, use of local materials,and earthquake-resistant construction features. Theobjective was not only to demonstrate the improve-ments in traditional building practices, but also togenerate confidence among the residents about theuse of stone and its by-products for housing con-struction. Out of 528 buildings, 400 were con-structed in the two most affected districts, Latur andOsmanabad. The remaining model buildings wereconstructed in the districts of Solapur, Satara, andSangli. The model houses had different types ofplans and material options, e.g., stone masonry,stone-crete block masonry,hollow block masonry, andclay brick masonry (WorldBank, 1994a).

Contractors under the supervi-sion of PMU engineers con-structed most of the modelbuildings. The program wascompleted in the initial year ofthe program implementation.During the MEERP implemen-tation, model buildings wereused as offices for the PMUengineering staff, or as part ofcivic amenities and otherpublic buildings plannedunder MEERP. In fact, some ofthe highest ranking PMU fieldengineers made the model

houses their official residences during the MEERPimplementation since Indian government agenciesprovide official accommodations for their employ-ees. In the Latur district, for example, the superin-tending engineer (the topmost GOM engineeringposition at a district level), and his subordinates(five executive engineers and over 20 deputyengineers) all resided in the stone masonry modelhouses (Figure 36). This gesture was an effective“confidence-building” tool for the many beneficia-ries who feared stone construction. The modelhouses incorporated seismic features and werehighlighted by bond beams that were painted inbright colors. Most model buildings were con-structed at exposed locations. In front of somebuildings there was a sign, written in the localMarathi language, which described the salientseismic features of that particular building.

Pilot Demonstration StrategyObservation■ In order to get the Reconstruction, Repair and

Strengthening Program (RRSP) under way, theGOM supported the construction of demonstrationhouses in some of the villages. This was intendedto help motivate other beneficiaries to participate.

Figure 36 A model house used as a residence of a PMUDeputy Engineer in Latur.


DiscussionIn the first year of the post-earthquake rebuildingproject, the PMU field staff found it very hard tolaunch the RRSP. The relevant GOM policy decisionsregarding the RRSP and the key features of thetechnology packages offered in the program weresummarized in a brochure that was printed in a locallanguage in August 1994 and circulated to themajority of beneficiaries. Senior GOM administrativeofficers at the district level toured the villages andconducted meetings to familiarize the beneficiarieswith the RRSP.

In spite of all those efforts, the GOM field engineersfound it difficult to get the rebuilding under way. Thebeneficiaries did not understand the time frame of therehabilitation nor the need to mobilize the materialsand labor required for the construction. In somevillages, none of the beneficiaries initially agreed tosign the consent for the RRSP, which was an essentialinitial step in the RRSP implementation procedure.

In an effort to facilitate implementation of the RRSP,the PMU decided to launch the Pilot DemonstrationProject in October 1994. The project was intended toassist the beneficiaries in the construction or strength-ening of at least one house in each village. The PMUengineering staff, particularly the JEs, assisted thebeneficiaries in their efforts to procure the materialsand artisans for the construction. Some JEs evenaccompanied beneficiaries to the material marketsand assisted them in the selection of building materi-als, e.g., cement, steel, sand, and bricks. Later, the JEsmobilized the construction labor (masons) andoriented them regarding earthquake-resistant stoneand brick masonry construction. The project provedto be very successful, and according to the projectchief engineer, it represented a major breakthroughin the RRSP implementation (Momin, 1998).

Pilot Strengthening Program (PSP)Observations■ The GOM used the rebuilding effort as an

opportunity to provide information on improvedconstruction practices, including the seismicstrengthening of undamaged structures in partsof the state not affected by the 1993 earthquake,by launching the Pilot Strengthening Program(PSP) and providing financial and technical

assistance to 5,000 owners of undamagedvulnerable houses in 13 districts of Maharashtra.Although the underlying idea behind this pilotprogram—to demonstrate seismic strengtheningtechnology using traditional rural houses—wasvery innovative and positive, the program wasdifficult for the GOM to manage simultaneouslywith the post-earthquake rebuilding effort. Thekey lesson learned in this pilot program is that theimplementing agency needs to plan the programvery carefully and ensure maximum technicalassistance and supervision in the implementationphase. In a first-time application of such aprogram, with a population that did not haveany previous exposure to earthquake-resistantconstruction practices, a single poorly retrofittedhouse might set a wrong example to others.

■ Strengthening of 46 undamaged public buildings(mainly schools) managed by the PMU wasincluded in the program at a later stage. Thisprogram was more successful in terms of qualityof technical solutions and construction, and itwas easier to implement than the strengtheningprogram for private buildings.

DiscussionThere are more than 2.5 million traditionally builtstone masonry buildings in the high-risk seismiczones in Maharashtra. In view of the completedestruction of or damage to more than 230,000houses in the 1993 earthquake, and the documentedvulnerability of this type of construction in earth-quakes elsewhere in the world, the GOM and theWorld Bank agreed to launch the PSP for 5,000private buildings scattered throughout 13 districts inthe state.

The GOM prepared the strategy document on thePSP (GOM, 1996a), which contained the policy forthe program implementation, including the criteriafor the selection of beneficiaries. According to thestrategy document, the houses that were supposedto be selected for the PSP were similar to thoseaffected by the 1993 earthquake (mainlyunreinforced stone masonry dwellings with interiortimber frames and heavy earthen overlays atop theroofs), with similar seismic strengthening featuresrequired. In order to ensure the demonstrationeffect, the GOM policy required that the housesincluded in the PSP be located in the central village


of each district. Implementation of the program waspartially sponsored by the GOM grant assistance,and partially self-supported by the homeowners.

According to the strategy document (GOM, 1996a),information dissemination was considered to be acritical aspect of the PSP. Information about programpolicy and the implementation methodology was tobe made available to the target communities viavarious media, including newspapers, radio, TV,mobile video units, street plays, and governmentpublications. The GOM prepared illustrative pam-phlets and posters on seismic strengthening tech-nologies in the local Marathi language and circu-lated them in the villages included in the PSP(Figure 37). In order to highlight the seismic fea-tures, special plaques describing the strengtheningwork that had been done were to be custom madefor each house.

The initial plan was to implement this programthrough the PMU. However, due to the large areacovered by the PSP (13 districts totaling approxi-mately 146,000 sq. km [56,370 sq. mi.]) and the factthat this program was scheduled for implementation

Figure 37A beneficiaryof the PSPshowing aknee-braceused toretrofit atimber framestructure.

simultaneously with the rebuilding of 230,000earthquake-damaged houses, the GOM believed thatit would be very difficult to stretch the PMUworkload any farther. (The PMU engineering cellhad a staff of over 900.) Therefore, the GOMdecided to implement this program through thelocal district governments—Zilla Parishads (ZPs) andtheir engineering units. However, it seems that thechoice of implementing agency in this case was notideal. ZP engineering units had several othersimultaneous assignments related to rural develop-ment programs in the districts. As a result, ZPengineers assigned to this program were not able toprovide adequate guidance and supervision to thePSP beneficiaries. By and large, the ZP engineersdid not have any previous exposure to earthquake-resistant construction technology. They were offeredtraining by the national seismic consultants to thePMU. The foreign seismic consultant to the PMUprepared technical guidelines for the PSP (GOM,1995a) that were circulated to the ZP engineers thatoutlined the vulnerable types of traditional housesand the corresponding seismic strengtheningprovisions (Figure 38).


Figure 38 The buildings selected for the PSP were located at the central locations in a village(building at the back, left). This village was a “market place” in the Ahmadnagar district, and aweekly market was taking place.

The program was implemented in the period fromMay 1996 to June 1998. It may be concluded that theimplementation of this program was not very success-ful, especially compared to the in-situ rebuildingprogram. With few exceptions, a majority of the PSPbeneficiaries carried out the construction. However,in many cases new construction, either extending theexisting house area or rebuilding on the existingplinth, was carried out instead of strengthening.Unfortunately, very few cases of houses strengthenedin accordance with the GOM technical guidelineswere observed during the authors’ field visits. Themain causes for inadequate quality of programimplementation appear to be the inadequate techni-cal assistance offered by the ZP engineers to thebeneficiaries and artisans, and the failure of the ZP asthe program implementing agency to launch aneffective information dissemination campaign. Due tothe lack of information provided in the early stages of

the program, the beneficiaries had no faith in tradi-tional construction methods and did not want toinvest funds in such buildings.

It is noteworthy that in 1997 the GOM and the WorldBank decided to expand the PSP to include thestrengthening of undamaged public buildings in sixdistricts of Maharashtra (the same districts included inthe PSP of private buildings). This program wasimplemented by the PMU. In total, 46 public build-ings (mainly school buildings) were strengthened.The buildings selected were of typical rural construc-tion, mainly unreinforced brick or stone masonrybuildings, 10 to 40 years old without any seismicfeatures. Strengthening of these buildings was doneusing the same technology as for the PSP involvingprivate buildings (Figure 39). The program wasimplemented in 10 months by a team consisting of anexecutive engineer, three deputy engineers and


several JEs of the PMU (in Latur). The program wassuccessful primarily for two reasons: 1.) the engineer-ing team was enthusiastic and had four years ofsimilar experience working in MEERP (i.e. in-siturehabilitation of earthquake-damaged buildings), and2.) the engineers were in control of the programimplementation, since the construction carried out bythe contractors was based on documents prepared bythe engineers. As this program was launched in thefourth year of the MEERP implementation, most ofthe other project components were almost com-pleted, enabling the PMU engineers to provideadequate field supervision in the construction phase.

Figure 39 Students and teachers in front of a retrofitted elementary school building inthe Latur district. The concrete bands are dark stripes near the top of the buildings.

Base IsolationDemonstration ProjectObservation■ To demonstrate effective strategies for earthquake

mitigation that could be used in future construc-tion of important facilities in Maharashtra, theGOM decided to construct two base-isolateddemonstration buildings close to the epicenter ofthe 1993 earthquake. Public buildings of masonryconstruction typical for the rural area wereselected. Although rubber bearings had to beimported for this application, the entire construc-tion was carried out using local artisan skills andtools and was supervised by the PMU engineers.This was the first reported application of baseisolation technology in India and one of the veryfew applications in rural areas worldwide.


Figure 40 Ahalf portion ofthe Killari schoolbuilding wasconstructed as abase-isolatedstructure (underconstruction inforeground),and the twinbuilding (brickmasonrybuilding in thebackground)was constructedin aconventionalmanner.

DiscussionBase isolation is an advanced technology for seismicprotection of building structures and their contents.The concept is that base isolation provides flexibilityin building structures by means of rubber bearingsinstalled underneath the superstructure at the plinthlevel, thereby “isolating” the structure from damag-ing ground shaking effects. This technology isespecially effective in providing seismic protectionto rigid building structures (e.g., load-bearingmasonry and concrete buildings). It is noteworthythat, by and large, fatalities in major earthquakes(especially in developing countries) have occurreddue to the collapse of rigid low- and medium-riselow-strength masonry buildings.

Even though the concept of base isolation is simple,several decades of development and sporadicapplications were required for it to gain wideracceptance in the worldwide engineering commu-nity. Effectiveness of this technology was confirmedin the 1994 Northridge and 1995 Kobe (Japan)earthquakes.

Most of the applications of base isolation technologyto date have occurred in industrialized countries,like Japan, the U.S., and New Zealand. Amongdeveloping countries, China appears to be at theforefront in the number of design applications

(following a United Nations Industrial DevelopmentOrganization (UNIDO)-sponsored project completedin 1994); some design applications have alsooccurred in Armenia (sponsored by a World Bankproject after the 1988 earthquake), Indonesia (aUNIDO-sponsored project), Chile, Mexico, andRussia. A major research effort in the area of baseisolation has taken place in India in the last 20years, especially at the Department of EarthquakeEngineering, University of Roorkee, with the firstPh.D. thesis on the subject completed in 1978(Qamaruddin, 1978). The research has mainlyconcentrated on seismic protection of masonrybuildings using simple and low-cost base isolationsystems based on the friction concept (Figure 40).

Advanced analytical studies related to the mechanicsof base isolation and the seismic response ofisolated systems were also carried out in severalacademic centers in India, especially at the IndianInstitute of Technology, Delhi and Powai (Mumbai),and the University of Roorkee. Unfortunately, nopractical construction application of this technologyin India has been reported to date.

In November 1997, the World Bank and the GOMdecided to use the post-earthquake rebuilding projectas an opportunity to demonstrate the practical applica-tion of base isolation technology for the first time in


India. Two base isolation demonstration buildingswere constructed at the new (relocated) Killari village,very close to the 1993 earthquake epicenter (approxi-mately 10 km [6.2 mi.] away). Two public buildingswere selected for this project—one a school buildingand the other a part of the shopping complex. Thebuildings selected were single-story with rectangularplans, made of brick masonry wall construction andconcrete roof slab and constructed with the seismicfeatures recommended by Indian seismic standards.Cement mortar was used in construction, and rein-forced concrete ring beams were constructed at thelintel level of the buildings. Vertical reinforcement wasnot provided. In general, the same technical specifica-tions and design features that were used in otherpublic buildings constructed in the relocation villagesunder MEERP were used here.

Figure 41 Installation of isolation bearings carried out manually using local artisansunder the supervision of PMU engineers.

The only additional construction features specific tothe isolated structures were concrete ring beamsinstalled below and above the isolation bearings(Figures 41 and 42). The isolation bearings weremade of natural rubber reinforced with steel shims. Abolted connection was used to attach the bearings tothe adjoining structures. Installation of the bearingswas completed in February 1999. The installation wascarried out manually, and the construction was doneby the local construction work force (masons) withexperience in masonry construction and without anyprevious exposure to complex engineering projects.The supervision during the construction was pro-vided by the PMU engineers assisted by the foreignseismic consultant to the PMU.


Figure 42 Construction of an upper concrete ring beam above the isolators for aschool building in Killari village.

Because the seismic provisions for base isolatedstructures are not included in the current seismiccode in India, design of the structures followed theseismic provisions for base isolated buildings of the1994 Uniform Building Code (UBC, 1994). Seismicityof the site was found to correspond to the UBC1994 Seismic Zones 2B or 3. Since this was the firstapplication of base isolation technology in India, thebearings were imported from the United States.Rubber bearings with mechanical characteristicssuitable for base isolation applications are notpresently available on the Indian market. However,since India is a producer of natural rubber, therequired facilities might be developed in the future,once this technology gains wider acceptance bygovernment agencies and the private sector.

To demonstrate to the general population theeffectiveness of base isolation technology in theabsence of seismic instrumentation, both demonstra-tion buildings were split into two identical portions(in terms of design and dimensions). One wasconstructed in a conventional manner, while theother was equipped with isolation bearings. It isexpected that the isolated buildings will remainundamaged even in a major earthquake that mightaffect the area in the future (the 1993 earthquakewas an intraplate earthquake characterized by anunpredictable recurrence pattern), whereas theconventional half of the buildings are expected toexperience structural and nonstructural damage.However, collapse is not expected to occur (com-patible with the performance objectives of thecurrent Indian seismic code).

Documents

Housing Reconstruction - nicee.org