8130 Field Report - Designing a world of hope | EMI 2 Project No. 8130 16 7-May-2007 It is well documented that brick masonry (and to a lesser extent, concrete masonry) expands and

FIELD REPORT

for

Samaritan’s Purse

Klieng Cot Aron Village Shelter Project

Banda Aceh, Sumatra Indonesia

5th May 2007

Prepared by: M. Coffey R. Koeniger

eMi2 Project No. 8130 16 7-May-2007

It is well documented that brick masonry (and to a lesser extent, concrete masonry) expands and contracts in the wall over time and temperature variations. Without any material in the wall capable of resisting these small tensile forces (caused by contraction), or alternatively, without any joints (free relief surfaces) to limit the length of the wall section subject to these internal expansion / contraction forces, the wall will crack. Cracks essentially relieve these expansion / contraction stresses and they develop because the bricks themselves, the mortar, and the plaster have little to zero ability to resist tensile stresses. This means that cracks due to such expansion / contraction are quite natural in un-reinforced masonry (such as that used at the Cot Aron Shelter project and most other buildings in Banda Aceh / Indonesia / India) and can be expected to occur. Provided they are not in a position causing a compromise in structural strength, these cracks are not dangerous – though at times they may be aesthetically unpleasing.

5.3 TYPICAL CRACKING OF LARGER FRONT GABLE FACE

Mr. Tarigan stated that the lack of rebar in a concrete section provided at the top edge of this brick masonry gable, or the lack of such a member is a main cause of this cracking. It is true that if such a section were provided in all cases it would tend to limit or reduce the severity of the type and nature of cracking observed at this location, but the omission of the reinforced concrete section (or sufficient rebar in the same) would not tend to cause cracking. The main cause for cracking in the larger front gable face of the Cot Aron Shelter units is believed to be related to the design of the ring beam supporting element. In reviewing the situation with Mr. Tarigan, he stated that in most cases the size of this ring beam was actually reduced in construction (he estimated 10x10cm as opposed to the 15x15 cm shown on the plans) and maintained that if the carpenters had followed the design on the plans, the situation would be fine. It was unclear whether he realized that the ring beam documented on the plans was not meant for this application at all and that a new design was called for. However, it is apparent in Mr. Tarigan’s statements about the gable design change for the Lamasan Shelter project that he was uncomfortable with this unsupported ring beam situation. As it so happens, the ring beam is just capable of supporting the basic dead load imposed by the brick gable end. The ring beam described on the plans is 15x15cm [5.9”x5.9”] with 4-12mm bars – two top, two bottom. Considering this beam having a concrete fracture strength (fc’) of 13.8MPa [2,000 psi], and an actual reinforcement cover depth of 1”, and an actual size of 10x10cm [3.9”x3.9”], the nominal moment strength of the beam is 29.6 in*kip. Considering the triangular un-factored dead load on the beam, the required moment strength is around 25.5 in*kip. Thus, the actual as-built capacity exceeds the actual as-built demand by around 16%. These values do not take into account any load factors or strength reduction factors (i.e. safety factors) which would mandatory considerations in proper a design. Were the requirements of a proper design considered, the factored demand moment of the beam exceeds the allowable moment capacity by around 34%. These values are mainly dependant upon the design assumptions for the actual constructed size of beam, steel size and arrangement, and the concrete fracture strength – all of which could be quite variable across the project considering the construction conditions. This ring beam is also not deep (stiff) enough to prevent deflections beyond the range tolerable for a brick masonry wall with plaster – especially considering the manner of construction and speed of loading. Also, compared to the masonry wall which supports the other half of the gable, the stiffness of the beam is very small. Both the former and the latter of these conditions (overly flexible beam or dramatic change in the stiffness of the supporting element) would lead to cracking consistent with that observed in the Cot Aron units. To complicate and compound matters further, the veranda end of the ring beam under the gable is supported by another beam which is likewise ill-suited and under-designed for the purpose. That beam, in its turn, has also likely deflected and in some cases cracking consistent in type and nature with this cause was observed in the corner at the top of the wall where this ring beam connects to the wall column.


This cracking may have been developing from the time of the gable’s completion, resulting from initial deflections of the ring beam due to pre-mature removal of the shoring and the quick loading of the ring beam. If this were the case, it would indicate that the cracks may not progress further. Alternatively (and considering the poor manner of design / construction), the cracking may be a progressive problem tending to worsen over time. Without observing the cracking over a period of time (to check progression), it is not possible to determine which of the two possibilities is the case. Also, it is important to point out that the majority of Shelter units did not show signs of cracking at the gable. This is likely due to an improvement of the conditions on any number of factors already mentioned which relate to the design / construction / loading of this ring beam. Steps should be taken to repair this type of cracking and remediate the structural situation at the beam.

5.4 TYPICAL CRACKING IN WALLS AND AT DOORS AND WINDOWS

The main cause for cracking in walls and at or around doors and windows of the Cot Aron Shelter unit (and at other SP Shelter projects) is believed to be the expansion and contraction stresses in the brick masonry walls. Compounding this cause are conditions of weak / poor mortar and plaster mixes and lack of curing for the plastered surfaces. Cracking resulting from expansion and contraction stresses usually tend to fail along the weakest (the smallest) cross-section of the wall, since that section will correspondingly have the weakest resistance to tension. This is consistent with the type and nature of wall, door & window cracks observed in the Cot Aron Shelter project. Better mix proportions and proper plaster curing would have limited or reduced the severity of this cracking. Wall reinforcing materials or the provision of control joints would have prevented or concealed it altogether. Above a certain threshold for crack width, steps should be taken to repair this type of cracking.

5.5 TYPICAL CRACKING IN TOILET WALL

The main cause for cracking in the toilet wall of the Cot Aron Shelter unit (and at other SP Shelter projects) is believed to be the lack of proper treatment in design and construction of the natural joint between the unit’s corner and the toilet wall end. The toilet wall brick masonry is most likely not (or very poorly at best) joined into the corner of the unit. That being the case, this toilet wall end does not appear to be significantly connected in any way to the corner of the unit, except by interior and exterior plastering. In this condition, cracking would be expected at this location. In many cases, a wood fascia board of the toilet roof penetrates the top of the toilet wall causing a stress concentration, providing an easy ‘head’ to the crack. The type and nature of the cracking observed at the toilet wall of the Cot Aron Shelter units is consistent with this cause. A connection along the top of the toilet wall to the unit corner could have contained this crack or prevented it from developing – or would have caused it to be one of the preceding category (4.4). Alternatively, a covered joint in the wall at this location would have prevented or concealed cracking at this location. Above a certain threshold for crack width, steps should be taken to repair this type of cracking.

6.0 OPTIONS FOR MITIGATION AND REPAIR

6.1 THE “ACCEPTABLE THRESHOLD” FOR CRACKING

Regardless of the many other complexities of the current situation at Cot Aron, it is clear from the survey of the Shelter project that there is some cracking which impacts only aesthetic appearance and some cracking which results from structural deficiencies and should be repaired. Much of this cracking is above the “acceptable threshold” which seems to be commonly tolerated in the area. The inspection survey conducted at and around Cot Aron indicates that the “acceptable threshold” total Shelter unit grade is around ‘3’ or ‘4’. Cracking at or below this level would be apparent in almost any building constructed of un-reinforced brick and plaster. The percentage of units in the survey which were graded above ‘4’ was 35%, as a simple average of series A, B & C. Across the entire Cot Aron Shelter project, it is


therefore estimated that about 80 units exhibit cracking above the “acceptable threshold”. In view of this, the following options for crack mitigation and repair are recommended:

6.2 GENERAL WALL CRACKING REPAIR

If any crack repair program is started at Cot Aron, the likely eventualities will be that each resident will carefully search their unit to point out even the slightest cracks (meaning those below the “acceptable threshold”) to receive repair and will be extremely conscious of which units receive repair and which do not. To try and avoid any complications which may arise from these eventualities, it is recommended that there be a general, three-level crack repair program, the levels of which are separated by a public physical test for crack width using shims of certain widths (objective) rather than visual inspection (subjective). See Appendix F. In light of the “acceptable threshold” of cracking, the shim widths proposed below are believed to be reasonable points at which to separate the repair levels and target cracks that should be repaired. Level 1: This level targets grade ‘1’ cracks and some minor grade ‘2’ cracks which are less than 1mm in width. Since all residents will likely demand attention for crack repair, it is likely that many times grade ‘1’ cracks or other minor cracks will be pointed out. Efforts should be taken to repeatedly kindly reassure the residents that the same are quite natural for these types of buildings, are not dangerous, can be seen in buildings throughout Aceh, and require no repair other than to paint over the crack to improve the appearance (if desired). In such cases, the level 2 shim can also be used to point out that the crack width is ‘too small’. Level 2: This level of crack repair targets grade ‘2’ and some grade ‘3’ cracks which are 1mm or greater (but <3mm) in width and would apply if a shim of a 1mm thickness fits into the crack. This would involve a simple crack patch of wall putty which would fill and cover the crack. The plaster along the crack line can be dug out to form a groove (~4-5cm width) so that any loose plaster is removed and a roughened surface is provided to receive the patch material. The crack and the wall surface should then be thoroughly cleaned out with a wire or hard bristle brush and the loose dust should be blown off. The putty should then be pointed into the crack as much as possible over the length of the crack. Finally, the groove should be filled in and smoothed over with the putty. Several units at Cot Aron have already been patched in a similar fashion. The residents would be capable of such repairs though, in view of the project history, it may be preferred to have a contractor conduct the work. Based on the number of units surveyed with grades of ‘2’ assigned on the elements of inspection, it is estimated that about 20-25% of the Shelter units (46-58 units) would receive repairs of this level. Level 3: This level of crack repair targets severe grade ‘3’ cracks which are 3mm or greater in width and would apply if a shim of a 3mm thickness fits into the crack. This crack repair is a more intensive repair aimed at reinforcing the wall with wire mesh. The plaster on either side of the crack would first be chipped away to a certain distance and the wall surface well cleaned. For applicable cracks at toilet wall locations, this distance should be 50cm on either side of the crack (~1m total width). For applicable cracks at other locations, this distance can be 25cm on either side of the crack (~0.5m total width). The crack length would then be filled by pointing either wall putty or a rich mortar into the crack. A single length of straightened wire mesh of width matching the prepared wall surface would then be fastened to the wall along the crack length such that a gap of ~5mm at least is preserved between the wall and the mesh. Finally, a rich plaster would then be applied to a thickness of at least ~10mm. The plaster should be worked onto the wall to fill the gap behind the mesh and its thickness should be sufficient to ensure it adequately contains and covers the wire mesh. Steps should be taken to ensure the plaster cures properly by regular wetting (or covering with a wetted cloth or sacking) to keep the patch in a damp condition for at least 7 days after it is applied. A contractor is recommended for this repair level. Based on the number of units surveyed with grades of ‘3’ assigned on the elements of inspection, it is estimated that about 5-10% of the Shelter units (12-23 units) would receive repairs of this level.


6.3 MITIGATION OF LARGER FRONT GABLE FACE ISSUES

Prior and in addition to the above wall cracking repair program, steps should be taken to improve the structural conditions at the larger front gable face location. This is due to the great number of unknowns regarding potential variations in the actual structural condition of the supporting ring beams, the possible propensity for further deflection, and concerns this area generates in terms of earthquake safety in its damaged state. If cracks are visible to the eye in the larger front gable face, it is recommended that the ring beam supporting the gable masonry be permanently unloaded halfway prior to crack repair according to the program outlined above, but at least of Level 2 repair. A certain portion of the larger front gable face masonry which is totally concealed by the smaller front veranda gable face should be permanently removed; see Appendix G. Permanently removing this load will significantly improve the structural situation at the ring beam both for stress and deflection considerations. Taking again the previous design assumptions and the actual as-built nominal moment capacity of this beam as 29.6 in*kip, the new loading condition actual as-built demand moment would be 19.4 in*kip (reduced from 25.5 in*kip). This means the actual beam capacity would exceed the actual loading demand by about 53% (improved from 16%). Were the requirements of a proper design considered in this improved situation, the factored demand moment of the beam would exceed the allowable moment capacity by only 2% (reduced from a 34% overstress). In order to do this work, either the ceiling from rooms below can be opened or one or two of the roof sheets removed temporarily. The portion to be removed constitutes the masonry starting at the mid-span of the ring beam / the peak of the veranda gable down to the roof eave point at the edge of the veranda. The wood purlins currently supported by this portion of the gable wall can be supported by wood posts cut to size which rest on the ring beam. The repair program should be phased such that at least 2-3 days are given between the removal of masonry and the application of either level of crack repair. A contractor is recommended for this work. Based on the number of units surveyed with grades of ‘2’ or ‘3’ assigned for the gable, it is estimated that about 30% of the Shelter units (69 units) would receive this remediation work.

6.4 SPECIFICATIONS FOR REPAIRS

• Steel shims of 1-2cm width and 1mm / 3mm thickness should be used for checking crack widths. A metalworking shop in Banda Aceh should be able to provide these from sheet stock or etc. without difficulty.

• Wall putty material (RJ Wall Putty) was locally available in one of the hardware shops near Cot Aron and looked suitable for Level 2 crack repairs. This putty was available in 1 kg and 5 kg tubs costing 12,000 (USD1.2) and 32,000 (USD3.2) Rupiah respectively. This material or equivalent may be able to be found in greater bulk / for less cost in Banda Aceh itself.

• Two types of steel wire mesh were locally available in hardware shops near Cot Aron as well as in Banda Aceh. The first was an 18x18mm grid of ~1mm wire and the second was a 12x12mm grid of ~0.5mm wire. Both were found in 1m wide rolls and both would be suitable for Level 3 crack repair, though the first type would be preferable. The lowest price found (in Banda Aceh) was 8,000 Rupiah (USD0.8) per square meter.

• Plaster / mortar mixed for crack repair should be proportioned at 1:4 or 5 (cement:sand). If the mix is proportioned per sack of cement, the volume of whatever practical instrument is used to measure out sand should be verified to determine how to preserve the proper mix ratio. If the SP wheelbarrow is used, this would mean one sack of cement to one wheelbarrow of sand. Mixing water should be controlled to not exceed 50-60% of the cement volume.

• Appendices F and G are drawings provided for use in conducting the cracking repairs outlined in the Field Report. These drawings can be reproduced and translated into Bahasa for use and other notes / instructions for the contractor may be added as necessary or as desired by SP.


7.0 CONCLUDING REMARKS

In the circumstances and pressures faced by Samaritan’s Purse in the execution of the Cot Aron Shelter project, a wide variety of events and decisions took place that have led to the challenges of the current situation. The cracking in the units, though primarily natural in its causes, has become more extensive than normal and in many cases exceeds the local “acceptable threshold” for cracking. This is due mostly to the poor techniques and quality of construction. This is very apparent when considering units built by contractors from other NGOs. The cracking at the unit’s front gable is largely the result of the use of an inappropriately designed / constructed beam to support the gable masonry. This situation came about through the use of plans which contained a flaw, an un-thorough or missing review of the unit’s structure after design changes, and structural oversights at the beginning of construction for the project. It is hoped that the information, analysis and recommendations provided in this Field Report will be helpful as SP decides how to proceed in a repair program for the project. It has been our intention to make such a program as simple, low-cost, and practically effective as possible, while remaining transparent in its criteria of operation. This is especially important since a cracking repair program will likely set a precedent for other villages and NGO Shelter projects in the area. It is hoped the program outlined will help to bring about a satisfactory and fair result for all parties concerned. We are thankful to have been given the opportunity to be of service to Samaritan’s Purse in consulting on the Cot Aron Shelter project and hope that EMI / eMi

2 will be able to continue to

be of assistance to Samaritan’s Purse, save in happier circumstances.


8.0 APPENDICES

8.1 APPENDIX A – SHELTER UNIT PLANS



8.2 APPENDIX B – AS-BUILT STRUCTURAL FRAMING PLAN


8.3 APPENDIX C – INTERVIEW WITH VILLAGE OFFICALS

Head of Cot Aron Village: Interview 24.04.07 Bahasa: “Kuchek” means head of village Head of Cot Aron Village: Zulkifli, S. Sos Community Representative: Suwardi, S. PdT. Secretary: Isaknder Former (resigned) Head of Cot Aron Village: Zakariah

1) Is the HoV elected? Yes 2) For how long is the term? 5 yr term, currently a temporary 1-year term. Election is in

2008 for the next full term. Zulkifli was elected in 2007 after Zakariah resigned in Nov 2006. Zulkifli is running again for office.

3) Does the HoV receive any salary or compensation for his position in the village? No, the position is voluntary.

4) What is the occupation of the HoV, where does he work? Government of Indonesia position in Cooperative Development and Finance under Governor of Banda Aceh.

5) Was there any compensation to families affected by the Tsunami of 26 Dec 2004 by the Govt. of Indonesia? No, but some household items were provided. Capital was given to businesses to restart.

6) What was the population before the Tsunami in Cot Aron? 300. After? 500 (between S.P. and Mamamia.)

7) Land use in Cot Aron prior to Tsunami? There were larger family homes and some wood or semi-permanent homes. Houses before Tsunami had foundations too. After Tsunami, families split into 2-3 units between several cottages.

8) How was property allotted by the govt. after the Tsunami? It was not overseen by the govt. Villagers worked it out with each other to settle the boundaries.

9) Did the HoV help to supervise construction of the homes? Supervision of the home construction was only by homeowner and the S.P. appointed person, the HoV did not personally oversee other villager’s homes being constructed

10) How were carpenters found for the construction? They were not provided by S.P. so each homeowner was responsible to find their own. As one house was completed, the carpenter would move on to the next house. There were not enough carpenters for each house project. There were approx. 20 head carpenters working in the village post-tsunami and usually 3-4 helpers/unskilled labor working on a house with one carpenter.

11) Who paid the carpenters? Samaritan’s Purse. S.P. Paid 3.5-3.7m Rupiya in an agreement between the homeowner and the carpenter. Many homeowners made additional deals with the carpenters to pay them additional funds above the agreed to rate given by S.P. because as the construction progressed, the carpenters asked for more money. Nearer to the end of the project, carpenters were typically asking 6.5-7m Rupiya per house. The homeowners complained to S.P. and the rates were increased by S.P. to make up the difference. S.P. reimbursed all homeowners for the difference in cost from the original agreement and the additional rates charged by the carpenters. Govt. of Indonesia set the rate for carpenters and it increased during the reconstruction of the homes in Cot Aron. Currently, a carpenter for a BRR house costs 11m Rupiah.

12) When were cracks in the homes first brought to the attention of the HoV? First reported by homeowners in July 2006, Suwardi reported to Governor, copied BRR and Head of Dist. (See copy of faxed letter sent from Banda Aceh to Medan internally by SP) A village meeting was held and the people of the village communicated verbally to the representative from the BRR the cracking happening in their homes. Mr. Suwardi also then went to the Governor on behalf of the people to voice complaint.

13) Are other people in other villages complaining to other NGOs about cracking in their homes? Yes, the BRR homes have cracking (Suwardi) and BRR is fixing the damages. Matthew asked then to see the BRR homes and Suwardi agreed to take him there. On 25.04.2007, Matthew and Kristiantos went to see the BRR homes. Sarasi (NGO) built 135 homes and the BRR paid out 15m Rupiya per house as a


compensation? Current quoted cost for building a home like the one in Cot Aron is 45m Rupiya.

14) What has caused the cracking in the homes in Cot Aron? HoV: each house is different. Saltwater used in construction. Salty sand. Brick is bad because it is from outside and SP bought the cheaper, low-grade bricks, not the local ones. Sometimes carpenter did not use rebar in the gable. SP paid lower rates, so carpenters rushed. Homes built in the end of the project are better because the carpenters slowed down and used more materials: rebar and cement and better brick and sand.

15) How many houses have the village leaders looked at? Almost all houses cracked, if not, then wood doors warped and split because unseasoned wood was used. All door hardware does not work. The HoV cannot measure how many homes have cracks, but we (Matthew and Ryan) are free to look for ourselves. HoV says ~230 houses need fixing. Suwardi took 152 photos of houses with cracks and submitted to BRR. Suwardi lost the copies of the photos they took and submitted to BRR in Feb 2007. Suwardi says they cannot ask for the copies of photos from the BRR because then they may not be able to get SP to fix the homes. (??) (Zulkifli stepped out to call someone on mobile phone)

16) How are they going to fix the cracks? We don’t know. A consultant contractor gave them a report (see photos from meeting) in which a letter was written and an estimate given for repairing all 230+ homes. Consultant contractor actually just signed the document and did not look at any of the homes. This was required by the BRR because they have stated that a village committee cannot submit a proposal for fixing their own homes, it must be done by a contractor. So, the HoV wrote/prepared the report and then had the contractor sign and stamp it so that it could be submitted to the BRR.

17) Why are there not cracks found in the carpenters houses? Because carpenter owns brick factory so he sold the SP bricks and then used his own superior quality bricks in his home. Bricks made in Cot Aron are all good. Also, had more time/care in making.

18) How many houses have only plaster cracks? Cannot guess.


8.4 APPENDIX D – SHELTER PROJECT SURVEY



8.5 APPENDIX E – BRICK TEST

8130 Klieng Cot Aron Field Brick Test (Coffey, 25.4.07) Purpose: To do a rudimentary and public demonstration for the people to check whether local bricks were indeed of a vastly better quality than bricks from outside Cot Aron. Bricks: 10 bricks were collected at random from a leftover pile found on the ground reportedly brought in from somewhere in Aceh. These bricks had been transported to Cot Aron and were lying on the ground for an unknown period of time. 10 bricks were collected from a local brick factory in Cot Aron. These bricks were hand picked from the kiln by the proprietor. Tests:

a. Water test: to determine initial density of bricks (2 bricks from each source) b. Drop test: to give an indication of relative durability / brittleness of bricks (8 bricks

from each source) Water test: A bucket was borrowed from one of the homes and filled with water. The water level was measured as well as the upper diameter of the water surface. Two bricks from the local source were placed in the bucket and the new, higher water depth was immediately measured. The same procedure was performed for two bricks of the outside source. The size of the two outside bricks were approximately the same as the size of the two local bricks. This experiment showed that the two outside bricks tested were denser than the local brick. The local brick was found to weigh approximately 0.31 kg [0.71 lb] per brick with a density of 495 kg / cubic meter [30.9 lb / cubic foot]. The outside brick was found to weigh approximately 0.42 kg [0.93 lb] per brick with a density of 499 kg / cubic meter [31.2 lb / cubic foot]. Drop test: A pre-cast concrete element was borrowed from a pile to use as a backdrop for the test. This element was stood on end on a concrete floor surface (an old building foundation washed away in the tsunami) and marks were made on the face of the element at 2’, 3’, 4’ from the floor surface. The floor was in public view near the main road through Cot Aron, and there were at least five to six local witnesses for this test. Two bricks from each source were selected for drop from a 4’ height, four bricks from each source were selected for drop from a 3’ height, and the final two bricks were reserved for ascending drops starting from 2’. The bricks were hand dropped to strike the floor evenly on the bottom (widest aspect) of the brick. This experiment showed that the eight outside bricks tested were tougher / less brittle than the local brick. From 4’, one of two bricks survived the first drop from each source. On the second drop from 4’ for the surviving bricks, the local brick failed and the outside brick chipped. From 3’, only one of four local bricks survived the first drop, while three of four outside bricks survived the first drop. On the second drop from 3’, the local brick chipped while two outside bricks failed and one outside brick chipped. From 2’, both bricks of each source survived the first drop, one of each source survived the second drop (from 3’), and the same again chipped in the third drop (from 4’). Conclusion: This rudimentary experiment was conducted more for the public benefit (though not purposely advertised or dramatic) to spread rumors about the qualities of local vs. outside brick than for its value as statistical / scientific data. It was claimed that the local brick was ‘all good’ while the outside brick was ‘sometimes good, but mostly bad’. Though the outside brick tested better, one may reasonably suspect the statistical reality is that the quality of clay brick made / fired / delivered by the same processes will be approximately the same regardless of the source.


8.6 APPENDIX F – CRACKING REPAIR


APPENDIX F – CRACKING REPAIR (REVERSE)

Crack Repair Program

Level I Minor crack, no repair required. The resident can cover it by painting.

Level II 1. Break away plaster from wall along

the crack to form a groove 4-5cm in width.

2. Thoroughly clean crack in groove with a wire brush and remove dust.

3. Fill in crack along its length with putty by pointing the putty into the crack.

4. Fill in groove and patch with putty along crack length.

Level III 1. Remove plaster from wall on either side of crack for:

a. 50cm at toilet wall b. 25cm at other locations

2. Thoroughly clean crack and wall with a wire brush and remove dust. 3. Fill in crack along its length with putty by pointing the putty into the crack. 4. Cut a single length of wire mesh to apply to the wall and attach with nails to keep

a gap of 5mm between wire mesh and wall. 5. Apply plaster to wall, keeping mesh contained and covered by plaster. 6. Keep plaster patch damp for 7 days after application.

Does 1mm shim

fit into crack?

Does 3mm shim

fit into crack?

No

No

Yes

Yes


8.7 APPENDIX G – GABLE REPAIR


TABLE OF CONTENTS

1.0 INTRODUCTION 3

1.1 SAMARITAN’S PURSE – INDONESIA OVERVIEW 3 1.2 TEAM PERSONNEL 3 1.3 SCOPE OF WORK 3 1.4 METHODS AND SOURCES OF RESEARCH 4

2.0 COT ARON PROJECT BACKGROUND 4

2.1 GENERAL TIMEL INE 4 2.2 PROJECT PERSONNEL TIMELINE 5 2.3 PROJECT EXECUTION METHOD 5 2.4 PLANS USED FOR SHELTER PROJECTS 5 2.5 CURRENT SITUATION AT COT ARON 7

3.0 DOCUMENTATION OF CRACKING 8

3.1 THE COT ARON SHELTER PROJECT SURVEY 8 3.2 CRACKING AT THE COT ARON SHELTER PROJECT 9 3.3 CRACKING AT OTHER NEARBY SP SHELTER PROJECTS 11 3.4 CRACKING AT OTHER NGO’S SHELTER PROJECTS 11

4.0 ANALYSIS OF CRACKING 11

4.1 REVIEW OF SUGGESTED CAUSES OF CRACKING 11 4.1 .1 EARTHQUAKES AND TREMORS 11 4.1 .2 LOW-QUALITY MATERIALS 12 4.1 .3 USE OF SALTWATER IN MIXING 12 4.1 .4 SHELTER UNIT SETTLING 12 4.1 .5 WORKMANSHIP 13 4.2 SHELTER UNIT CONSTRUCTION CONDITIONS 13 4.2 .1 MIX PROPORTIONS 14 4.2 .2 SHORING OF GABLE RING BEAM AND CURING 14 4.2 .3 UN-REINFORCED BRICK MASONRY CONSTRUCTION 15 4.3 TYPICAL CRACKING OF LARGER FRONT GABLE FACE 15 4.4 TYPICAL CRACKING IN WALLS AND AT DOORS AND WINDOWS 16 4.5 TYPICAL CRACKING IN TOILET WALL 16

5.0 OPTIONS FOR MIT IGATION AND REPAIR 16

5.1 THE “ACCEPTABLE THRESHOLD” FOR CRACKING 16 5.2 GENERAL WALL CRACKING REPAIR 17 5.3 MITIGATION OF LARGER FRONT GABLE FACE ISSUES 18 5.4 SPECIFICATIONS FOR REPAIRS 18

6.0 CONCLUDING REMARKS 18

7.0 APPENDICES 19

7.1 APPENDIX A 19 7.2 APPENDIX B 201 7.3 APPENDIX C 212 7.4 APPENDIX D 24 7.5 APPENDIX E 26 7.6 APPENDIX F 27 7.7 APPENDIX G 28


1.0 INTRODUCTION

1.1 SAM ARITAN’S PURSE – INDONESIA OVERVIEW

Samaritan’s Purse is a Christian International Aid and Relief organization (NGO) which is engaged in various projects around the world. In response to the earthquake and tsunami which occurred near Sumatra, Indonesia on 26 December, 2004, Samaritan’s Purse (SP) began relief operations in Sumatra in early 2005. Currently there are several field bases which execute SP projects (such as Shelter projects) and one central office currently located in Medan. A field base in Banda Aceh, one of the areas worst affected by the tsunami, was the center for the Shelter projects discussed in this Field Report. The Shelter project at Klieng Cot Aron (pronounced Chot Ah-roan) is the focus of this Field Report. The SP central office is headed by a Country Director (CD) and field bases are headed by a Base Manager. Additionally, the Shelter projects are overseen by a Shelter Manager. At the time of the Shelter projects mentioned, a Mr. Paul Gimson was the SP CD in Medan. In the same period the Banda Aceh base manager position was filled by five different people for varying amounts of time. During the majority of the Cot Aron Shelter project, a Mr. Randall Knutson worked as both the Shelter and Base Manager at Banda Aceh. With one exception, none of these expatriate field staff are still with SP.

1.2 TEAM PERSONNEL

Engineering Ministries International (EMI) is a non-profit Christian development organization whose vision is to mobilize design professionals to provide design assistance to selected non-profit organizations in developing nations. The design professionals work free-of-charge and their work is coordinated through one of five EMI offices located in Colorado, India (eMi

2),

Guatemala, Canada or Uganda. EMI is the most experienced cross cultural and professional non-profit design service available in the world. This was the first time eMi

2 has partnered with Samaritan’s Purse and the first visit to

Indonesia by personnel from eMi2. The field consultation team included:

Matthew Coffey eMi

2 Director, Structural Engineer

Ryan Koeniger eMi2 Project Development, Architect

1.3 SCOPE OF WORK

As per the contract agreement with Samaritan’s Purse, the team was in Medan, Banda Aceh, and Calang, Sumatra, Indonesia from April 19-29, 2007 for the purpose of completing the following:

• Collect and review available information regarding the Cot Aron Shelter project;

• Interview available SP personnel and others who were connected to the Cot Aron Shelter project;

• Observe and document field conditions of the housing units of the Cot Aron project and other nearby SP projects and similar projects by others;

• Produce Field Report on the Cot Aron project. The purpose of this Field Report is to document the field notes and findings of the eMi

2 team

in Medan / Banda Aceh / Calang, address various concerns or related issues, and to advise SP of options and methods for project repair.


1.4 METHODS AND SOURCES OF RESEARCH

Unfortunately in the process of closing the Cot Aron project and the Banda Aceh base, many of the paper records such as receipts, delivery records, and progress reports were thrown away or lost. Though the project computer in the Banda Aceh office unfortunately broke before EMI arrived, the data was recovered and reviewed. This Field Report is based on available documentation EMI found in the SP Medan and Banda Aceh offices and also interviews with available SP personnel previously connected to the Shelter projects in question. Where inconsistencies between interviews occurred and no documentation was found, the Field Report presents a generalized consensus.

19

th, 20

th April – Medan, Sumatra:

• Met with Mr. Jeff Sandwick (SP Indonesia Country Director), Mr. Mike Brewton (SP Indonesia Deputy Director) and Mr. J.P. Mellis, (SP Regional Director)

• Reviewed plans used for the Shelter projects in Aceh

• Reviewed the various project files collected from the Banda Aceh project office 20

th – 26

th April – Banda Aceh, Sumatra:

• Visited the Lamasan and Durung Shelter projects with Mrs. Sarida Harianja (SP Medan HR) and Mr. Kristanto (SP Banda Aceh (BA) base)

• Inspected first 21 constructed, last 21 constructed, random 20, and worst damaged (10 nos.) Shelter units constructed at Cot Aron

• Inspected total of 20 Shelter units at random by four different NGOs in Cot Aron area

• Inspected 10 Shelter units at random constructed at Lamasan

• Inspected soil conditions at Cot Aron and Lamasan village areas

• Performed a small, rudimentary brick test (comparing local brick vs. outside brick)

• Interviewed Mr. Kristanto (former SP BA base Shelter construction supervisor) and Cot Aron village headmen (both current and former) and community representative

26

th – 27

th April – Calang / Laguen, Sumatra:

• Interviewed Mr. Atas Tamba (former SP BA base procurement), Ms. Grace Sebastian (former SP BA base administration), and Mr. Reyes Rodriguez (former SP BA base Shelter manager - Durung)

27

th – 29

th April – Medan, Sumatra:

• Interviewed Mr. Ikut Tarigan (former SP BA base Shelter coordinator)

• Reviewed available paper and computer files from BA base

2.0 COT ARON PROJECT BACKGROUND

2.1 GENERAL TIMELINE

In early 2005 after the Banda Aceh SP base was established, the first SP Shelter project (~35 units) was started in the village of Durung, about 30 minutes outside the city. In mid-2005 the Cot Aron Shelter project (~230 units) was started in Klieng Cot Aron village, about 15 minutes outside Banda Aceh. In early 2006 a Shelter project (~130 units) was started in Lamasan village close by to Cot Aron. A fourth, smaller Shelter project (~20 units) was started in mid to late 2006 in a village called Mataea, also near Banda Aceh. These Shelter projects overlapped with each other such that start and end dates were not clear, but it seems most of the Shelter units in Cot Aron were completed by March or April in 2006. After completing the final project in the fall of 2006, the Banda Aceh base was closed and the staff (with the exception of Mr. Kristanto) moved to other SP projects in Indonesia. The Banda Aceh base currently serves as logistics support for communications and official work with the Bureau of Rehabilitation and Reconstruction (BRR) which is a regional (Aceh Besar District) Government office established to regulate and coordinate NGO reconstruction activities.


2.2 PROJECT PERSONNEL TIMELINE

After the completion of the Durung project, the Cot Aron project was started administratively sometime in June-July 2005 under Mr. Reyes Rodriguez (construction background, Shelter construction experience) continuing from Durung as the Shelter manager. Community meetings were held in Cot Aron and a warehouse was constructed (finished end of June 05) and materials were collected to begin the project. Shortly after this Mr. Rodriguez left SP and Banda Aceh in August 05 (he later returned and worked on different projects). The Shelter unit construction at Cot Aron was begun sometime in August-September 2005 in the absence of a Shelter manager or SP BA base staff with a technical or construction background. In November 2005, Mr. Ikut Tarigan, an Indonesian civil engineer, was hired and joined the base staff as Shelter coordinator. At this time approximately 10-15 Shelter units in the Cot Aron project were either completed or in various stages of construction. Mr. Tarigan was with the BA base until its closure. The SP Banda Aceh base was being managed by a Mr. Joel Crocker – from unknown date until end June 05, followed by a Ms. Tamara – for a two week period, followed by a Ms. Jackie Blevins – from July? to December 05, followed by Mr. Randall Knutson from November 05 to unknown date in 06 (presumed after completion of the Cot Aron project), followed by a Mr. Paul Hickok from mid-June 06 to closure of the BA base. It is not clear how exactly these Base Managers overlapped or when or if there were gaps in the base management.

2.3 PROJECT EXECUTION METHOD

The Cot Aron project started at the invitation (18.04.05) from certain members of the Cot Aron village government for SP to construct permanent homes in their village. This was approved by the regional Government and the project was subject to BRR regulations. Unlike many other NGO’s Shelter projects, the SP Shelter projects did not use a contractor for construction but opted rather to involve the beneficiary (homeowner) in the construction of their own house with the help of a skilled workman (head carpenter). Since the earthquake and tsunami destroyed the existing homes and boundaries, the village land in Cot Aron was freshly subdivided into plots for the beneficiaries by the regional Government (or an NGO appointed / approved by the regional Government). The beneficiaries were to be those families whose homes were destroyed in the tsunami – one home per family as they existed prior to the tsunami. The Cot Aron Shelter project was executed on the basis of two main contract documents: 1. Memorandum of Understanding (MoU) between the regional Government and SP

(23.01.06 – formalized near project completion) which outlines the parties, the number of units to be constructed, and some basic terms for the Shelter project.

2. Contract documenting terms of payment for skilled labor between SP, the beneficiary (homeowner), and a head carpenter. This contract was made on a per-unit basis prior to the start of construction of the Shelter unit and the payment was arranged in several stages (i.e. foundation complete, walls complete, etc.) to retain payment leverage for each unit. To account for wage inflation / other NGO’s precedents / market value for a carpenter, this contract was adjusted at least once during the Cot Aron project.

Additionally, multiple SP supervisors managed the stock, delivery, and use of the materials for the Shelter project. Documentation included receipts, stock tallies, progress reports, delivery slips, & etc. These supervisors did not typically have a technical / construction background. Each Shelter unit was built by a team of workmen consisting of a head carpenter, two or three carpenter’s helpers plus a number of laborers and other helpers. In the Cot Aron project, it seemed 20 to as many as 50 such teams were constructing Shelter units. Head carpenters came from a variety of places (i.e. Aceh, Medan) to work on the project and a few head carpenters were from Cot Aron. Each unit was built in an average of four to eight weeks and the total cost per unit was in the range of 35,000,000 Rupiah (USD3,500).

2.4 PLANS USED FOR SHELTER PROJECTS

The plans used for Cot Aron seem to have originally come from the regional Government of Aceh Besar / the BRR (or an Aceh engineering company commissioned by these parties);


see Appendix A. It seems these plans were first given to the SP Banda Aceh base by a former SP staff member working on a separate project in Melavo at the time. The plans were intended for use in the SP Durung Shelter project which was in the beginning stages at the

time. These plans describe a permanent Shelter unit of approximately 36 square meters [390 square feet]. The unit was originally designed such that the gable ends of the unit would be closed off with wood cladding attached to the wood trusses which were the main roof framing of the unit. The walls were to be of masonry with a reinforced concrete section at the plinth and top of wall (ring beam) levels. Additionally, reinforced concrete columns are called for at corners and certain wall intersections on the plan. An outside toilet connected to the back of the unit is described without much detail. On the front aspect of the house, a small gable covers the entry veranda, supported by two wood posts. Behind the entry veranda gable there is a jog in the exterior wall which creates a complication as to where the front-most roof truss should be placed. The plans do not describe a roof framing plan and it is unclear as to how the front truss end (at the veranda side) is to be supported. Presumably, this was an oversight / error in the original plans that carpenters / builders needed to handle upon reaching this stage in the construction. This front truss is supported along the top of the front-most wall for half its length and, in many cases, the veranda truss end was simply left hanging in the finished building. [Note the unsupported condition of the left bearing end of the front truss in the left-hand picture of Figure 1 below.] The Durung Shelter project used these plans with little or no deviation, aside from the use of plywood framed interior partition walls, and the relocation of a wall column to support a truss end.

Figure 1 At or shortly before the beginning of construction of the Cot Aron Shelter units, a decision was taken by SP to construct the Shelter units with masonry gable ends and interior walls. Apparently this change was due to both the high cost of wood and Government restrictions on illegal tree cutting which reduced wood availability. In the case of the interior walls, the BRR requested the use of masonry presumably for fire safety reasons. The switch from wood to brick was also aesthetically desirable from the Cot Aron beneficiary’s point of view. Unfortunately the same flaw in the original plans at the larger front gable face still existed and created a compound problem since the larger front gable face was now to be in brick masonry rather than simply a wood truss with wood cladding. The actual constructed height of the gable peak from the ring beam (~1.8 m [~6 feet]) is less than that shown on the plans (~2.4 m [~8 feet]). [Note the straight larger front gable face in plastered brick masonry in Figure 2 below.] At the start of the project, in the absence of an SP BA base manager, Shelter manager, or other staff with a technical / construction background or any other apparent technical / construction guidance, it seems the first units constructed at Cot Aron relied on the expertise of the head carpenter along with help from the SP construction supervisors. The practice at this gable condition was to extend the ring beam along the front wall to connect into the perpendicular ring beam on the edge of the veranda; see the as-built structural framing plan in Appendix B. At or after the arrival of Mr. Tarigan to the Cot Aron project (10-15 units underway) there seems to have been no change in the (by then) status quo method of framing the larger front gable. Mr. Tarigan confirmed that the design of the ring beam on the original plans was used in this situation.


Figure 2 Figure 3 At the beginning of the Lamasan project the plans for the Shelter unit were revised from that used at Cot Aron in that the larger front gable face was not kept straight, but jogged along with the unit’s walls below. [Note the jog in the larger front gable face in Figure 3 above.] This eliminated the need for extending the ring beam as described above for the Cot Aron units. According to Mr. Tarigan, (who suggested and received approval within SP for this change) this was done because he did not like the condition of “not having a wall beneath the ring beam” at the larger front gable face situation in the Cot Aron units and also to save the material / cost of constructing multiple ring beams. The change was not made at Cot Aron partway through because of the need to keep the design and appearance of the units all the same to prevent against problems with the beneficiaries who would likely take offense at any inequity. At Mataea, a design similar to Lamasan was implemented. In general, these plans were reviewed and discussed in the beginning of each project with the head carpenters and worked through during the construction of the first few units. After that point, the carpenters had more or less absorbed the patterns / processes of the unit’s construction and would rarely need to refer to the plans.

2.5 CURRENT SITUATION AT COT ARON

The current head of village (HOV) at Cot Aron along with the community representative (both are official positions, appointed by a majority vote in the village) said the first time cracking was observed in the units at Cot Aron was in the middle of 2006. Multiple claims were made as to what was done at that time or how thoroughly the 234 units at Cot Aron were inspected by these officials. The HOV sent a letter to the BRR, copying SP (3.1.07) requesting their assistance to renovate the units constructed by SP stating that they have been heavily damaged due to the use of low-quality materials and to the frequent earthquakes / tremors in the region. The village “community” also prepared a proposal for submittal to the BRR (8.1.07) describing the work which needed to be done in the Cot Aron units. Since the BRR only accepts proposals such as these from contractors, the village “community” and HOV used letterhead of a nearby contractor to write the document (along with a full cost estimate). The contractor came and visited the project, looking at one (1) of the units (the HOV’s house), and then signed the proposal. Unfortunately, the BRR replied that they were unable to take on work of such tremendous scope as the proposal requested and referred them to SP. The proposal calls for 15,000,000 Rupiah (USD1,500) per unit for 213 units bringing the grand total to over 3 billion Rupiah (USD300,000). This estimate per unit appears to be for re-building of certain sections of the units, not for repair work. The HOV mentioned units (~130) built by another NGO (Sarasi Indonesia) in a nearby village which had cracking that the BRR funded repair for. After visiting the village with Mr. Kristanto and talking with some Sarasi beneficiaries there, it appears that the BRR made an arrangement for a two-stage payment of 15 million for each Sarasi unit. The reason for this, however, appears to be to maintain equity


because Sarasi units were in a state ‘less complete’ than BRR units which are being built in the same village. One Sarasi unit was found which had some crack patching work done, but it seems that most of the people used the money (6 million Rupiah in stage 1) primarily to install floor tile / paint / install ceilings / build a kitchen as an addition / and etc. In February 2007 it seems that the Cot Aron village officials looked at a certain number of units exhibiting the worst cracking and took photos (152 nos.) of the damage and submitted the same to a certain Mr. Gerwanza Putra at the BRR (possibly to encourage them to take up the previously submitted proposal). Mr. Kristanto was asked to confirm this the day following the interview and Mr. Putra stated that he did receive these pictures, though he was unable to confirm when he received them or find the pictures to allow Mr. Kristanto to make copies for EMI. It was also reported that an article about problems with the Cot Aron project recently (in the last month) appeared in an Aceh newspaper. It seems that SP came to know of the article via official channels. At the time, Mr. Kristanto was unable to find the article in Banda Aceh newspapers and it was not clear if anyone at SP had ever seen / read this article. It is clear that there is cracking in the units in Cot Aron that exceeds the “acceptable threshold”. After walking around the village for several days inspecting over 70 units particularly and seeing many more casually, an interview with the village officials was conducted; see EMI’s interview notes in Appendix C. In the end, after working through various stories and establishing various points and explaining the purpose of EMI’s visit on behalf of SP, it appeared that the HOV and village officials were willing to accept that there were cracks in a certain, limited number of units that were ‘bad’ and needed to be fixed and that there are other cracks in a greater number of units which are natural and can be found in any building, including other NGO’s Shelter projects, Government buildings, and the village mosque.

3.0 DOCUMENTATION OF CRACKING

Kindly refer to Appendix D for the full data and details of the Shelter project survey

3.1 THE COT ARON SHELTER PROJECT SURVEY

There are four main categories of cracking observed at Cot Aron which constituted the ‘elements of inspection’ for the survey conducted in the Shelter project. These categories are: Gable, Wall, Door & Window, and Toilet. Using a visual grading scale, these elements were inspected at a total of 72 units within the Cot Aron Shelter project. These units were selected in four series:

‘A’ – 21 units constructed first (according to Mr. Kristanto) within Cot Aron ‘B’ – 21 units constructed last (according to Mr. Kristanto) within Cot Aron ‘C’ – 20 units selected randomly from beneficiary name list (excluding units already selected in series A & B) within Cot Aron ‘D’ – 10 units exhibiting the worst conditions of cracking (according to Mr. Kristanto & available beneficiaries at the time of the survey) within Cot Aron

The visual grading scale consisted of the following:

‘0’ – no cracking visible ‘1’ – hairline cracking of plaster or ‘crazing’ of plaster ‘2’ – visible cracking (<1.5mm gap) of plaster and wall, limited crack path, limited depth ‘3’ – visible cracking (>1.5mm gap) of plaster and wall, full crack path, full depth

The method of inspection was according to the following: Mr. Kristanto was responsible to guide the survey to the correct homes selected for the survey and determine the names of the residents / the number of the unit being inspected. Mr. Koeniger recorded the inspection data and the resident's name / house number and any comments on the inspection. Mr. Coffey performed a visual inspection of each Shelter unit by observing the condition of the 'elements of inspection' and then accessing one grade to each element per unit. The visual inspection involved observing the condition of the front gable, walking around the perimeter of the unit to observe the condition of the walls and at doors & windows, observing the condition of the


toilet wall, measuring the width of certain wide cracks, and in some cases (where the resident wished to show us certain cracks), observing the condition of the unit from inside.

3.2 CRACKING AT THE COT ARON SHELTER PROJECT

On an absolute scale, cracking damage observed at the Cot Aron project can generally be classified as ‘slight damage’. This category would include plaster and wall cracks with negligible to limited structural performance implications. To easily summarize the cracking inspection, the grades of the inspection elements were added together for each Shelter unit to give an idea of the overall condition of the unit. A perfect unit grade (no cracking) is zero while the lowest grade (most cracking) is 12 – a grade of ‘3’ on all four of the elements of inspection. The average total unit grade given to the Cot Aron Shelter survey of 72 units was 4.9 out of 12.0. The worst average total unit grade occurred in series D and was found to be 7.9 out of 12.0. Removing series D from consideration, the average total unit grade is 3.9 out of 12.0. There does not appear to any distinguishable pattern considering the unit grades in the A, B, & C series, though the C series had the best average unit grade. The cracking at the elements of inspection of the SP Shelter units at Cot Aron is described and summarized as follows:

A. Gable cracking [Figure 4]: Occurring in the larger front gable face of the building. At their fullest expression, the cracks are of full wall depth and proceed from the peak or top edges near the peak of the gable triangle, around the edges of some vent blocks, and down through the wall body to approximately the midpoint of the bottom leg of the gable triangle. The cracking tends to be of a wider width near the top surface of the gable, narrowing as it proceeds downwards. The average grade given to gable cracking over all 72 inspected units is 1.2 out of 3.0. The worst average grade of cracking occurred in series D and was found to be 2.3 out of 3.0. Removing series D from consideration, the average grade given for gable cracking is 0.9 out of 3.0. 14 units out of the 72 (19%) were assigned a grade of ‘2’ and 12 units out of the 72 (17%) were assigned a grade of ‘3’.

B. Wall cracking [Figure 5]: Occurring in areas of

plain masonry wall without penetrations. At their fullest expression, the cracks are of full wall depth and proceed from the top edge of a long plain wall section down to the bottom of the wall at the plinth beam. The cracking tends to be of an even width and the crack path tends to be a fairly straight line. Also, it typically occurs in the center of the wall section’s length. In several cases this type of cracking was also present at the rear gable of the unit. In a few cases this type of cracking occurred horizontally along a plain wall. [The picture shows a patched wall crack so that crack is visible to the camera]

The average grade given to wall cracking over all 72 inspected units is 1.3 out of 3.0. The worst average grade of cracking occurred in series D and was found to be 2.3 out of 3.0. Removing

Figure 4

Figure 5


series D from consideration, the average grade given for wall cracking is 1.0 out of 3.0. 14 units out of the 72 (19%) were assigned a grade of ‘2’ and 7 units out of the 72 (10%) were assigned a grade of ‘3’.

C. Door & Window cracking [Figure 6]:

Occurring at or around doors and windows or other wall penetrations. At their fullest expression, the cracks are of full wall depth and proceed from the corner of the penetration in a straight or diagonal direction to the nearest relief surface such as another door or window edge, the top or bottom of the wall, or a corner. The cracking tends to be of an even width and the crack path tends to be a fairly straight line. [The picture is a close up of a crack at the bottom edge of a window] The average grade given to door & window cracking over all 72 units is 1.3 out of 3.0. The worst average grade of cracking occurred in series D and was found to be 2.2 out of 3.0. Removing series D from consideration, the average grade given for wall cracking is 1.0 out of 3.0. 21 units out of the 72 (29%) were assigned a grade of ‘2’ and 3 units out of the 72 (4%) were assigned a grade of ‘3’.

D. Toilet wall cracking [Figure 7]:

Occurring in the joint between the wall of the toilet and the corner of the unit. At their fullest expression, the cracks are of full wall depth and proceed from the top edge (or from the bottom edge of the toilet roof fascia board which often penetrates into this wall) of the sloping toilet wall where it abuts the corner of the unit downward to the bottom of the wall at the plinth beam. The cracking tends to be of a wider width near the top of the toilet wall, narrowing as it proceeds downwards. The crack path tends to be straight and vertical. In one case, the crack in this situation appeared to be clearly linked with some settling of the toilet area. The average grade given to toilet wall cracking over all 72 units is 1.3 out of 3.0. The worst average grade of cracking occurred in series A and was found to be 2.0 out of 3.0. Removing series A from consideration, the average grade given for toilet wall cracking is 1.0 out of 3.0. 13 units out of the 72 (18%) were assigned a grade of ‘2’ and 14 units out of the 72 (19%) were assigned a grade of ‘3’.

E. Other cracking: In one case, the cracking of the unit appeared to be clearly linked with

some settling at a wall corner. Diagonal cracks were located at the tops and bottom of the wall at the corner in question. Additionally, the column at the corner in question was several inches out of plumb.

Figure 6

Figure 7


3.3 CRACKING AT OTHER NEARBY SP SHELTER PROJECTS

The SP Durung Shelter project was visited and a brief, random inspection tour was performed at the SP Lamasan Shelter project. As mentioned previously, the front gable condition at these projects was different than at Cot Aron and was not a significant cracking issue at these projects. Wall type cracks and door & window type cracks were present to a comparable degree with those at Cot Aron. The toilet wall condition was very similar at these projects and the cracking observed was to a similar degree as at Cot Aron. The average total unit grade given to the Lamasan Shelter series of 10 units was 3.5 out of 12.0. 2 units out of the 10 were assigned a grade of ‘2’ and 1 unit out of the 10 was assigned a grade of ‘3’ for wall cracking. 4 units out of the 10 were assigned a grade of ‘2’ and 1 unit out of the 10 was assigned a grade of ‘3’ for door & window cracking.

3.4 CRACKING AT OTHER NGO’S SHELTER PROJECTS

Four different NGO Shelter projects were visited in the Cot Aron area and brief, random inspection tours were performed at each project. Once again, the gable condition specific to the SP Cot Aron units does not apply to other NGO’s work. Similarly, since different designs were used for the toilet, the toilet wall condition was not comparable with the SP units. Wall and door & window type cracking was observed, but usually to a lesser degree than that observed in the SP Cot Aron project. About half of these Shelter projects were of more recent construction than the SP Cot Aron project and all of them were built by contractors. The average total unit grade given to other NGO Shelter survey of 20 units was 2.8 out of 9.0. 3 units out of the 20 were assigned a grade of ‘2’ and 2 units out of the 20 were assigned a grade of ‘3’ for wall cracking. 7 units out of the 20 were assigned a grade of ‘2’ and no units were graded ‘3’ for door & window cracking.

4.0 ANALYSIS OF CRACKING

4.1 REVIEW OF SUGGESTED CAUSES OF CRACKING

Various causes have been supposed by different parties to account for the cracking observed at the Cot Aron SP Shelter project. The most pertinent are considered individually below.

4.1.1 EARTHQUAKES AND TREMORS

Both the village officials and individuals at SP have pointed to earthquakes as a main cause of cracking. Sumatra and specifically the Aceh region have experienced a number of earthquakes and tremors of various intensities in the period between the 26 December earthquake and tsunami until date. They continue to frequently occur in the region and one even occurred (27 Apr) in the Banda Aceh region during EMI’s visit. Though ground motions from these earthquakes (the ones which are felt at Cot Aron) would tend to compound or grow existing cracking, the types of cracking observed at the Cot Aron Shelter project are not consistent with the wall shear cracking patterns and wall corner and intersection separation cracking that one would expect to observe were the damage a result of an earthquake(s). It has also not been claimed (or otherwise substantiated) that the cracking which has occurred appeared suddenly or as a direct result of any specific earthquake event. Finally, it is also true that these short units of lightweight construction are a small risk (low hazard) for damage or collapse by minor or indirect (distant epicenter) earthquake ground motions such that one would not expect them to sustain much damage from the same.


4.1.2 LOW-QUALITY MATERIALS

The village officials point to low-quality materials (meaning here the use of ‘outside’ materials – primarily brick) as a main cause of cracking. A small, rudimentary public experiment was performed at Cot Aron to compare local vs. outside brick ‘quality’. The methods and results of these experiments can be seen in Appendix E. The experiments showed that random bricks sampled from local (Cot Aron) and outside sources failed similarly (though the actual result of the tests was that the outside brick was of ‘better quality’). It was apparent that the reason SP was unable to purchase bricks locally in Cot Aron was due to the cost (which was opportunistic at best, approaching extortion at worst) quoted to SP / other NGOs. It appeared the local officials were still feeling a bit down about the loss of this potential profit, which connects quite neatly to the claims of superior local brick quality. Mr. Tarigan stated that the beneficiary did not tend take care of the material ‘packets’ given to them by SP as instructed (i.e. covering the brick from rain, storage of wood, cement & rebar, etc.). It is reasonable to say that material quality is more or less constant across most NGO’s Shelter projects since the same material sources / brick making methods were being used are basically the same (at least those in the Cot Aron / Banda Aceh area). Undoubtedly, the quality of materials used is low to fair, but this is found across the board in the region and does not shed much light on cracking specific to the Cot Aron Shelter project.

4.1.3 USE OF SALTWATER IN MIXING

Both the village officials and individuals at SP have pointed to the use of saltwater in concrete / mortar / plaster as a cause of cracking. It appears that, due to the location of Cot Aron to the coast, saltwater or salty groundwater was used to mix concrete / mortar / plaster and this is blamed for the low strength of the same. According to research on the topic, however, the use of saltwater in mixing actually tends to enhance the short-term strength gain / setting time of concrete / mortar / plaster. Long-term strength is affected such that mixes using saltwater are seen to attain strengths around 85% of comparable mixes using freshwater. With the collection of other construction method / quality control issues apparent in the Cot Aron Shelter project, it does not seem reasonable to point to long-term strength loss of such a minor degree as a result of the use of salty water in mixing as a significant cause of cracking.

4.1.4 SHELTER UNIT SETTLING

Individuals at SP have pointed to settling as a cause of cracking. The soil conditions at the unit foundation level at Cot Aron were found to be a type of well-consolidated brown silty sand. Mr. Kristanto stated that this type of soil was relatively consistently found throughout the Cot Aron Shelter project. At Lamasan, the soil conditions at the unit foundation level were found to be a well-consolidated black silty or sandy clay with a significant amount of organic content. These soils were classified by simple visual and hand tests. The most critical soil bearing pressure of the Shelter unit is found at gable wall locations and is on the order of 17-19 KPa [350-400 psf], considering that the width of the foundation is 0.5 m [1.6 feet]. It is also the case that many of the units constructed at Cot Aron have been built partially on the foundations of the old (destroyed) home. This condition sets the stage for differential settling damage since the old foundation will have settled to a certain degree (depending on age) and would be less prone to settle along with the new foundation. It was also stated by former SP BA staff that in several cases the Shelter unit was built on a wet, swampy area which was cleared off and filled in. This condition would set the stage for damage from excessive settling of the unit. Though many situations of units in Cot Aron potential soil settling issues may be expected through the circumstances, cracking in units which appeared to be directly connected to settling was only seen in a few cases. The soil bearing pressure is


quite low and would not seem to indicate a settling risk for the soils encountered. Though foundation settling can be independent of this, it is also many times related to soil bearing pressure. Also, the type of soil encountered at Cot Aron does not indicate a predisposition to settling. Again, settling of the units may compound cracking or cause crack growth but it does not appear to be the main cause in the majority of the inspected units.

4.1.5 WORKMANSHIP

Individuals at SP have pointed to workmanship as one of the main causes of cracking. SP made a decision to conduct the Shelter projects via a community-based method. The idea was to involve the beneficiary in the construction / management / supervision of their unit’s construction. At the time of the Cot Aron project there were numerous (~100) NGOs working in the Banda Aceh area, many of them conducting Shelter or etc. projects. All of these NGOs were looking for materials / contractors / carpenters to execute these projects. As a result (and to the delight of any person able to hold a hammer), the market value for carpenters / skilled workmen kept increasing, keeping carpenters in high demand. It is generally agreed the SP contract for the head carpenter was low from the beginning, and the beneficiaries (unknown at the time to SP and against contract rules) had to negotiate with the carpenters and supply additional payment in order for the carpenters to take up the work. In the end, the contract payment by SP was increased by a third and even then many beneficiaries seem to have incurred an out-of-pocket loss to the carpenters. It was reported that many carpenters took up construction on multiple units at one time to increase their earnings. The average time of construction for a Shelter unit in Cot Aron was recalled to be 4-8 weeks. It was also a value for SP that the Shelter project be completed quickly to enable the beneficiaries to move in as quickly as possible. Though the power of payment for work was in hand, the amount of control SP had over the workmanship of the Shelter units is uncertain. Most former SP BA staff stated that the carpenter rushed the work and did not listen to the instructions / quality control measures given by SP managers / SP construction supervisors. These instructions / quality control measures are said to have included (at least) the following:

a. method of storing materials b. method of mixing of concrete / mortar / plaster c. method of casting reinforced concrete sections with rebar placement & cover d. method of preparing the foundation soils (apparently in the placement of

bottom gravel and sand layers) e. method of curing plaster (only Mr. Tarigan mentioned this)

Mr. Tarigan stated that some of the units he saw upon joining the Cot Aron project were of very poor quality and a certain number (presumably a few) were broken down and rebuilt, though this was very unpopular with the beneficiary and other beneficiaries in a similar situation refused to allow their units to be reconstructed in this way. After being unable to satisfactorily explain the superb condition of the Shelter units of the head carpenters living in Cot Aron (besides claiming that the head carpenters managed to use the ‘superior’ local materials), village officials appeared to only reluctantly accept workmanship as a potential cause of cracking. The village officials stated the cause for this was the contract under-payment of the carpenters by SP.

4.2 SHELTER UNIT CONSTRUCTION CONDITIONS

Several pertinent conditions or methods of construction deserve mention as they are believed to apply directly to the cracking occurring in the Shelter units. They are considered individually below:

Documents

8130 Field Report - Designing a world of hope | EMI 2 Project No. 8130 16 7-May-2007 It is well documented that brick masonry (and to a lesser extent, concrete masonry) expands and