Deveopment of a Building Inventory for Manhattan Region

Department of Civil Engineering
The City College of the City University of New York
Convent Avenue at 138th Street, NY 10031

List of Figures

List of Tables

Summary

A research team from the City University of New York (CUNY) has been collaborating with a team of researchers from Princeton University on the development of a building inventory for the Manhattan area, to be used as part of the NYCEM Earthquake Loss Estimation Study. The study is funded by the Federal Emergency Management Agency (FEMA) region II and is coordinated by the Multidisciplinary Center of Earthquke Engineering Research (MCEER). The herein reported study has been supported by funds provided by MCEER.

The objectives of the CUNY team have been two-fold: (1) to assist in the collection and compilation of field survey data; and (2) to verify and expand upon the analysis of this data. The work completed so far can be divided into four tasks:

Preliminary findings of the study indicate that: (i) significant differences exist between the structural characteristics of Manhattan buildings and those in the default HAZUS inventory; (ii) loss estimation results appear to be sensitive to the building inventory, earthquake magnitude, and soil conditions. These findings are in agreement with earlier results reported by the Princeton team. Recommendations for future research include: (i) analysis of additional regions in Manhattan; (ii) modification of HAZUS to incorporate special types of buildings such as pile-supported tall buildings in downtown Manhattan (below 14th street); (iii) evaluation of earthquake loss due to soil liquefaction; (iv) incorporation of bridges and other lifelines into the study.

1.0 Introduction

Significant interest has emerged in recent years about the seismic risk in the New York Metropolitan area (Berneutter et al 1984; Jacob and Turkstra 1989; Nordenson 1995). The reasons are: (1) earthquakes are not unkonwn in New York City (two moderate events with maximum Modified Mercalli (MM) intensity of VII have sruck the area in the last 250 years; (2) earthquakes of magnitude 7 have occurred in other places in the Eastern United States; (3) there is a huge concentration of assets in the area; (4) the civil infrastructure in New York City has not been designed for earthquke loads, so significant damage may be caused even by moderate shaking.

To mitigate the earthquake risks in the area, a New York City Consortium for Earthquake Loss Mitigation (NYCEM) has been recently formed under the coordination of the Multidisciplinary Center for Earthquake Engineering Research (MCEER) at SUNY-Buffalo. The Consortium encompasses several interested parties such as engineers, architects, seismologists, major stock holders, emergency response agencies, real estate agencies, insurance companies, academia, and others.

An earthquke loss estimation study for the New York City area is currently being conducted by the Consortium. The study is funded by the Federal Emergency Management Agency (FEMA) Region II and is coordinated by MCEER. The major tool for perforiming these loss esimations is HAZUS, a GIS-based computer platform which has been developed by the Federal Emergency Management Agency through the National Institute of Building Science (NIBS). As part of this study, a research team from the City University of New York (CUNY) has been assisting a team of researchers from Princeton University in the development of a building inventory for the Manhattan area. The inventory is used to feed HAZUS with region-specific information, in order to enhance loss estimation predictions. Preliminary results from these studies have already been reported (Nordenson et al 1999; Jacob 1999).

The objectives of the CUNY team have been two-fold: (1) to assist in the collection and compilation of field survey data; and (2) to verify and expand upon the analysis of these data. The work completed so far can be divided into four tasks:

2.0 Completed Phase

2.1 Training of Research Assistants on HAZUS
Two undergraduate research assistants (Walter Fish and Paul Spiteri) attended the week-long session on the HAZUS program organized by FEMA in June 1999. Following this training, HAZUS-97 was installed on a high-end Xeon Pentium PC at the City College of New York. The installation was recently upgraded to the latest version of the program, HAZUS-99. Programs ARCVIEW, MAPINFO and ACCESS were also installed to facilittate processing of inventory information.

2.2 Survey of Kips Bay and Rockefeller Center areas
Our collection of field survey data consisted of the entire Rockefeller Center census tract and a portion of the Kips Bay tract (Fig 1). The Rockefeller tract is made up of 5 city blocks in midtown Manhattan, stretching from 5th Avenue to 6th Avenue between 50th and 55th Streets. (The city blocks correspond to No’s 1266, 1268, 1269, 1270, and 1271 according to the Sanborn Maps). The portion of Kips Bay tract surveyed by the CUNY team is made up of 5 blocks on the Lower East Side, from 1st to 3rd Avenue between 29th and 34th Streets (blocks No 939, 910, 911, 912, 935). A uniform data collection sheet prepared by the Princeton team was used for all buildings, which was modeled to coincide with the required fields within HAZUS. Specifically, each building was described by 8 entries namely: Frame Type, Exterior Wall Type, Basement Grade, Construction Quality, Exterior Condition, Existing Damage, HAZUS Building Code, and Occupancy. A digital picture of each building within the tracts was also taken, using a digital camera provided by the Princeton team. A total of 271 buildings (103 and 168 in Kips Bay and Rockefeller Center Areas, respectively) were surveyed by the CUNY team. This information along with the digital pictures were transferred into electronic format for distribution within the teams.

2.3 Processing of survey data into an ACCESS Database and statistical analyses
The collected data was processed into an ACCESS database. The format used is compatible with the required fields within HAZUS. A copy of the developed database is provided in Appendix I. Preliminary statistical analyses of the survey data were performed, as shown in Figs 2-9 and in Tables 1-4. With reference to the Rockefeller census tract, Figures 2 and 3 compare graphically the building occupancies in the default HAZUS inventory versus the survey data. The corresponding data are listed in Table 1. The differences in the number of buildings (649 versus 103 in the HAZUS default and the actual data respectively) and the percentage of residential buildings (12.3% versus 53.4%) are evident. Of particular interest is the significant difference in the number of commercial buildings (465 versus 42) which comprise the major occupancy type in the tract. A comparison of the corresponding building types is shown in Figs 4 and 5, and in Table 2. According to the survey data only two building types exist in the tract: unreinforced masonry (57.3%) and steel (42.7%). In contrast, a much wider variety of buildings exists in the default inventory. As will be shown later on, this will lead to differences in the earthquake loss estimations.

Table 1. Comparison of Building Occupancy Type for the Rockefeller Census Tract.
GENERAL BUILDING OCCUPATION	HAZUS DEFAULT	ACTUAL
Commercial	465	42
Educational	1	2
Government	0	1
Religious	24	3
Agricultural	1	0
Residential	80	55
Industrial	78	0
Total	649	103

Table 2. Comparison of Building Type for the Rockefeller Census Tract
GENERAL BUILDING TYPE	HAZUS DEFAULT	ACTUAL
Wood	161	0
Steel	251	44
Reinforced Concrete	37	0
Precast Concrete	17	0
Reinforced Masonry	51	0
Unreinforced Masonry	132	59
Mobile Home	0	0
Total	649	103

Figures 6-9, and Tables 3-4 compare data from Kip's Bay census tract. The CUNY team surveyed approximately two thirds of the buildings in the tract. Since not all survey data were available to the CUNY team when this report was written, the listed numbers have been extrapolated from the surveyed data assuming that the surveyed portion was representative of all buildings in the tract. Accordingly the number of buildings in each category are multiplied by 1.5 [ = 1/(2/3)]. The differences in the overall number of buildings (458 versus 253 in the default inventory and the actual data, respectively) as well as in building types (see Figs 8 and 9) is evident.

Table 3. Comparison of Building Occupancy Type for Kip’s Bay Tract
GENERAL BUILDING OCCUPATION	HAZUS DEFAULT	ACTUAL
Commercial	58	8
Educational	2	3
Government	1	0
Religious	8	3
Agricultural	1	0
Residential	383	237
Industrial	5	2
Total	458	253

Table 4. Comparison of Building Type for the Kip’s Bay Census Tract
GENERAL BUILDING TYPE	HAZUS DEFAULT	ACTUAL
Wood	270	0
Steel	38	20
Reinforced Concrete	18	0
Precast Concrete	2	0
Reinforced Masonry	34	0
Unreinforced Masonry	96	233
Mobile Home	0	0
Total	458	253

2.4. Preliminary loss estimation analyses using HAZUS
Preliminary analyses of hazard data were performed using the Wall Street, Rockefeller Center, and Kips Bay census tracts. The Wall street tract was used primarily for comparisons with results reported by the Princeton team. In all analyses, HAZUS99 for ArcView was applied to evaluate the damages that would be induced by an earthquake at a fixed location. Three different earthquake magnitudes (5.0, 6.0 and 7.0) were used at a location of latitude 40.56° N and longitude 74.0° W, which coincides with the estimated 1884 historical epicenter.

2.4.1 Wall Street Census Tract
Selected results from analyses using the Wall street census tract are provided in Tables 5 and 6. Corresponding results reported by the Princeton team (Nordenson et al 1999) are also shown in the tables. A description of the census tract is provided in the aforementioned report. It is seen that the two sets of values are very similar; the discrepancies may possibly be attributed to the different versions of HAZUS (99 vs. 97 version) used in the two analyses. The differences in the analyses using the default and the modified inventories is also evident, and it is due to the significant differences in the building types, occupancy and square footage in the two inventories.

Table 5. Struct. damage costs (thousands of $) in Wall Street census tract for an earthquake at the 1884 epicenter. Comparison of two HAZUS analyses.
Soil Type	Bldg. Inv. Type	Magnitude	Princeton Value (HAZUS 97)	CUNY Value (HAZUS 99)
Default	Default	5.0	3,369	3,369
		6.0	41,249	45,337
		7.0	187,388	198,124
Modified	Default	5.0	1,701	1,703
		6.0	26,079	28,415
		7.0	151,426	156,737
Default	Modified	5.0	772	770
		6.0	82,287	81,118
		7.0	283,809	279,644
Modified	Modified	5.0	218	219
		6.0	46,296	49,575
		7.0	228,635	235,952

Table 6. Cost of total loss (thousands of $) in Wall Street census tract for an earthquake at the 1884 epicenter. Comparison of two HAZUS analyses.
Soil Type	Bldg. Inv. Type	Magnitude	Princeton Value (HAZUS 97)	CUNY Value (HAZUS 99)
Default	Default	5.0	59,322	61,478
		6.0	412,136	415,673
		7.0	1,713,500	1,722,941
Modified	Default	5.0	24,530	25,492
		6.0	260,632	265,728
		7.0	1,403,898	1,409,352
Default	Modified	5.0	13,735	13,011
		6.0	445,526	439,357
		7.0	1,651,316	1,640,718
Modified	Modified	5.0	4,068	4,120
		6.0	272,707	275,981
		7.0	1,307,527	1,318,176

2.4.2 Rockefeller Center and Kips Bay Census Tracts
With reference to the Rockefeller census tract, Tables 7 and 8 compare loss estimation predictions using the default and modified building inventories. For simplicity, default soil conditions were assumed in all analyses. Also tall buildings (above 15 stories) that contribute a large percentage of the overall square footage, yet they are expected not to develop damage in New York City-type earthquakes were excluded from the analyses. It is seen that with the modified inventory the earthquake losses are significantly smaller than those with the default inventory ($3.7 million versus $25 million in structural damage for an earthquake of magnitude M=6, respectively). The differences can be attributed primarily to the significantly smaller number of buildings in the modified inventory (see Tables 2 and 4). The trend is, understandably, stronger with increasing earthquake magnitude. Similar trends are observed with the Kips Bay census tract (Tables 9, 10).

Table 7. Struct. damage costs (thousands of dollars) for an earthquake of fixed location (1884 epicenter) in Rockefeller census tract.
Soil Type	Bldg. Inv. Type	Magnitude	CUNY Value (HAZUS 99)
Default	Default	5.0	1,305
		6.0	25,356
		7.0	80,425
Default	Modified	5.0	1,073
		6.0	3,753
		7.0	11,557

Table 8. Cost of total loss (thousands of dollars) for an earthquake of fixed location (1884 epicenter) in Rockefeller census tract.
Soil Type	Bldg. Inv. Type	Magnitude	CUNY Value (HAZUS 99)
Default	Default	5.0	33,772
		6.0	288,993
		7.0	855,809
Default	Modified	5.0	3,883
		6.0	35, 602
		7.0	107, 547

Table 9. Struct. damage costs (thousands of dollars) for an earthquake of fixed location (1884 epicenter) in Kip’s Bay census tract.
Soil Type	Bldg. Inv. Type	Magnitude	Princeton Value (HAZUS 97)	CUNY Value (HAZUS 99)
Default	Default	5.0	337	501
		6.0	4,414	9,513
		7.0	21,524	27,471
Default	Modified	5.0	47	214
		6.0	2,277	4,610
		7.0	12,952	13, 310

Table 10. Cost of total loss (thousands of dollars) for an earthquake of fixed location (1884 epicenter) in Kip’s Bay census tract.
Soil Type	Bldg. Inv. Type	Magnitude	Princeton Value (HAZUS 97)	CUNY Value (HAZUS 99)
Default	Default	5.0	4,402	16,647
		6.0	49,521	112, 068
		7.0	186,793	290,571
Default	Modified	5.0	813	5,296
		6.0	20,462	40,402
		7.0	85,509	104,439

3.0 Conclusions and Recommendations for Future Research

A detailed building survey in two census tracts in midtown Manhattan was performed, to be used as part of the NYCEM earthquake loss estimation stydy. Kips bay and Rockefeller center areas were surveyed and the results were processed on an ACCESS database. A statistical analysis of the data showed significant differences between the strutural and occupancy characteristics of the surveyed buildings as compared to those in the default HAZUS inventory. Preliminary loss estimation analyses using HAZUS 99 indicate that: (i) good aggreement exists between the predictions of HAZUS 97 and HAZUS 99; (ii) results appear to be sensitive to the building inventory, earthquake magnitude, and soil conditions. The latter finding is in agreement with earlier results reported by the Princeton team.

Recommendations for future reserch include: (i) analysis of additional cencus tracts in Manhattan; (ii) modification of HAZUS to incorporate special types of buildings such as pile-supported buildings in downtown Manhattan (below 14th street); (iii) evaluation of earthquake loss due to soil liquefaction; (iv) incorporation of bridges and other lifelines into the study.

4.0 Acknowledgements

Financial support for the present study was provided by the Multidisciplinary Center for Earthquake Engineering Research (MCEER). The Authors would like to acknowledge the help of the Princeton team (Guy Nordenson, George Deodatis and Michael Tantala) in providing maps, inventory forms, and digital equipment for the survey. The Princeton team also provided innventory information for the Wall Street census tract. The Authors would also like to acknowledge the roles of Andrea Dargush (MCEER), Bruce Swiren (FEMA), Dan O’Brien (NYSEMO), and Klaus Jacob (LDEO), in the study.

5.0 References

Federal Emergency Management Agency (FEMA), (1997). 1997 NEHRP Provisions, FEMA 222A

Jacob, K.H. and Turkstra, C.J., eds. (1989). Earthquake Hazards and the Design of Constructed Facilities in the Eastern United States, Annals of the New York Academy of Sciences, Vol. 558, NY, 1989.

Jacob, K. H. (1999). "Site Conditions Effecting Earthquake Loss Estimates for New York City", Technical Report Prepared for MCEER, May 1, 1998 to April 30, 1999, Lamont-Doherty Earth Observatory (LDEO) of Columbia University

National Institute of Building Sciences (NIBS) (1997). Earthquake Loss Estimation Methodology: HAZUS Technical Manual, Volume I, Prepared for the Federal Emergency Management Agency (FEMA), NIBS Document Number 5201

Nikolaou, A. (1998). "A GIS Platform for Earthquek Risk Analysis", Ph.D. Dissertation, State University of New York at Buffalo

Nordenson, G.J.P. et al,. (1987). Seismic Hazard Evaluation for New York City, NYACE Ad-hoc Seismology Liaison Committee, Final Report, New York

Nordenson, G. J. P., Deodatis, G., Tantala, M., Kumpf, A., (1999). "Earthquake Loss Estimation Study for the New York City Area", Technical Report prepared for MCEER, May 1, 1998 to April 30, 1999, Princeton University