S THE SOCIETY FOR EARTHQUAKE AND CIVIL E S E C E D ... · THE SOCIETY FOR EARTHQUAKE AND CIVIL ENGINEERING DYNAMICS NEWSLETTER Volume 19 No 2 March 2006 ... Page 10 Mallet Milne Lecture

ISSN 0967-859XTHE SOCIETY FOR EARTHQUAKE AND CIVIL

ENGINEERING DYNAMICS

NEWSLETTERVolume 19 No 2

March 2006

SECED NEWSLETTER - MARCH 2006 - Page 1

SE

S E C E DED

Contents

Page 1 Indian Ocean Tsunami Damage

Page 3 Scratching the Surface: TheGeology of Earthquakes

Page 6 The Estimation of Peak GroundAcceleration in the UK after anEarthquake

Page 9 Blast Intensity

Page 10 Mallet Milne Lecture andReception 2007

Page 11 Report on the SECED YoungEngineers Conference, March2005

Page 11 Seismic Design to Eurocode 8

Page 11 Forthcoming Events

Page 12 Notable Earthquakes October –December 2005

Indian Ocean Tsunami DamageDr Tizianna Rossetto (University College London) and Dr Chris

Browitt (University of Edinburgh) report on the causes of damageas noted by EEFIT’s field investigation team.

IntroductionThe tsunami on 26 December, 2004,was a rare event which causeddamage in 12 countries around theIndian Ocean (Fig 1) and resulted inthe deaths of over ¼ million of thenationals of more than 50 countries.This death toll from a natural disasteris on a par with that of the Tangshan,China, earthquake of 1976 and is onlysurpassed by the Shansi, China,earthquake, in 1556.

Following the tsunami, the lead authorjoined the UK Earthquake EngineeringField Investigation Team (EEFIT) in itsmissions to Sri Lanka and Thailand. Itspurpose was to study damage to thebuilt environment in relation to thedegree of inundation and quality of thestructures, and to report back findingsand lessons learnt to engineering anddisaster mitigation communities.

The study areas provided a largevariation in the degree of damageexperienced; from total devastation,

with 70% of the buildings collapsed, tolight impact on windows and shutters.Overall, damage was the result of thecombination of structural vulnerabilityand exposure to lateral forces,

To put the Indian Ocean tsunami in the context of natural disasters over thepast 50 years; together with hurricane Ivan, it lifted economic losses for2004 to over $100bn. This has been only exceeded in 1995 when the Kobeearthquake, itself, cost around $150 billion. Most importantly, it demonstratedthe continuing exponential increase in total (and insured) losses, worldwide,from natural disasters; a trend that could be reversed with better planningand redirection of resources to the preparation phase rather than purely torelief. To learn lessons for future planning and construction, a UK fieldinvestigation team of engineers (EEFIT) visited Sri Lanka and Thailand toobserve and record the damage done.

themselves influenced by shoretopography, bathymetry, vegetationand coral formations, as well as theheight of the tsunami waves. In Sri

Figure 1. Countries directly affected by the tsunami


Lanka, average water heights weremeasured at 4.5m, with peaks of 10m,and, in Thailand, at 6m and 11m,respectively. Water velocitiesestimated from video footage, at twosites, were 3-4 m/s and 6-8 m/s (note:only Olympic sprinters can match thehigher speeds).

Causes of DamageThe three main causes of damagewere identified as ground scour, debrisimpact and water inundation. Largehorizontal forces are associated withwater movement and turbulence, andare enhanced by the entrainment ofsediments. Hydrodynamic andhydrostatic pressures reached 30kN/m2.Windows and masonry panels had littlechance as they fail at around 15% and30% of this level, respectively. Debrisimpacts (eg cars, boats, trees) wereclearly the source of some damage,including to exposed structuralmembers, but it was not possible toascertain its contribution to the totalcollapse of buildings.

The main cause of damage was dueto wave impact and inundation,reducing in intensity with distance fromthe coastline and, critically, with thenumber of obstacles between abuilding and the sea (vegetation, otherbuildings, infrastructure). A significantmechanism, particularly for thebridges, was the influence of groundscour (erosion). In addition to changingthe coastal topography, ground scourwas seen to have caused the failureof buildings founded on pad footings,and of roads and railways where theywere underlain by sand. Scour damageoccurred with the incoming tsunamiand also on the backflow, the latterbeing the main cause of failure ofbridges surveyed in Sri Lanka. Theirabutments were rock armoured on theseaward side but not on the inlandside, permitting erosion by backflowwater down the river channel and thecollapse of the reinforced concretedecks supported between theabutments.

Low-rise masonry houses, whichformed the majority of the buildingstock in Sri Lanka, suffered severedamage where water heightsexceeded only 2 metres. Those withreinforced concrete (RC) frames faredbetter although poor roof connectionsand scour around footings, induceddamage.

Three to five-storey RC-framedbuildings, which included most of thehotels in Thailand and public buildingsin Sri Lanka, did not fail although theirwindows and infill masonry panels did.This allowed water inundation. Theseare engineered buildings with inherentearthquake resistance, although notspecifically designed to withstandearthquake shaking. In this study,EEFIT engineers recognised that thedesign criteria for saving lives in anearthquake (so that the building doesnot collapse) need to be furtherenhanced for life-safety in tsunamivulnerable regions. Here, waterinundation is the main threat to life.

Economic Resilience?In Thailand, the greatest economicimpact has been on tourism, but, withmost hotels being insured, full recoveryis expected within one to two seasons.Consequential losses in Sri Lanka aresignificant in the tourism and fishingindustries with failure of the latterhaving an important longer term impacton the health of a population dependingon fish for 60% of its animal protein.

For much of the built environmentrelating to international tourism, thelessons learnt should aid resilience tofuture events.

But for others, here is a question anda challenge:

With the now familiar allegationsfollowing disasters in Developingcountries, that generous aid funds areslow to reach the people directlyaffected, is it time for the financialsector to devise “affordable” insuranceschemes for fishermen and smallbusinesses, so that rapid restorationof their commerce can act as thebottom-up engine to speed recovery?

For further information, visitwww.eefit.org.uk

Figure 2. Damage to buildings in Khao Lak, Thailand (courtesy of EEFIT).


Scratching the Surface: TheGeology of Earthquakes

The issue of rational, scientificallydefendable source characterization isthe key to good practice in ProbabilisticSeismic Hazard Assessment (PSHA).Many of the criticisms of PSHA havestemmed from either inaccurate orpoorly constrained sourcecharacterization or from a lack ofunderstanding of and/or poorcommunication of the uncertainties inseismic source parameters. This oftenresults in confusion over the meaningof hazard values at a particular site.Developing a robust seismic sourcecharacterization requires as full anunderstanding of the tectonicenvironment as possible and aconsideration of all possible faultbehaviour.

There are three simple questions toask when characterising potentiallysignificant seismic sources:

� Where are they?� How large are they – or what is the

maximum magnitude of earthquakethat they can generate?

� How often do they generateearthquakes?

The relatively short duration of theinstrumental and historical seismicrecords pose significant problems forseismic hazard assessment. Is the 100year snapshot provided byinstrumental seismicity or the partialview of an incomplete historical recordsufficient to quantify seismic hazard?In the majority of cases, the answer isa resounding no! So, short of statisticalgymnastics and ill-advised temporalextrapolations, how can we betterunderstand the vagaries of theearthquake machine? Althoughresulting from processes occurring

kilometres beneath the Earth’s surface,the surface manifestations ofearthquakes, in particular surface faultrupture, offer us clues of prehistoricearthquake activity.

Using tectonic geomorphology, thesignature that an earthquake leaves onthe surface of the Earth, we can useour knowledge of recent, historicalearthquakes to look for indications ofearthquakes in the recent geologicalpast. This, in ideal environments,allows us to extend the seismicityrecord from the relatively shortdurations of the instrumental (102

years) and historical (103 years)records to something approaching thereturn period for large earthquakes onfaults with the lowest slip rates (104 –105 years).

The geologist, especially when armedwith a strong background in structuralgeology, Quaternary stratigraphy, and

geomorphology, has all the tools forunravelling the secrets of pastearthquakes. Using the Law ofUniformitarianism (simply stated, thepresent is the key to the past) and thesurface expression of knownearthquakes as a yardstick, we caninterpret the earthquake recordpreserved in recent geological strata.However, as with all things geological,the record is not always clear andalmost never entirely complete!However, by examination of thevarious types of geological signaturesof past earthquakes, we can begin toquantify the uncertainties arising fromthe interpretation of such data, andwhat impacts this can have on seismichazard assessment.

As with all geological investigations,the study of seismic sources begins ata regional level before focusing inincrementally to reach a site-specificlevel. After all, it does help to

Scratching the Surface: The Geology of Earthquakes

This well-attended gathering began with the presentation by Brian Baptie to Piroozan Aminossehe of the prize for theannual SECED British earthquake prediction competition. Students of form will be interested to know that the squarewhere the event in question occurred had already provided a previous winner.

Then the meeting was treated to a presentation by Clark Fenton of the Department of Civil & Environmental Engineering,Imperial College entitled “Scratching the surface: the geology of earthquakes”. Essentially, this was an extensively(and very well) illustrated lecture. However, the speaker has been good enough to provide the following text as arecord for those who were unable to be present.

David Mallard provides an account of the 30 November 2005 Meeting and Clark Fenton’s lecture.

Surface faulting along the Lavic Lake fault – after the Hector Mine earthquake, southern California, M 7.1, 16/10/99.


understand the tectonic environmentand what style of faults you should beexpecting before you start trenching!

The use of remote sensing imagery,including satellite images andstereoscopic aerial photographs,allows us to locate structures withtectonic geomorphology that indicatesrecent movement, thereby answeringthe first question of seismic sourcecharacterisation: Where are theseismic sources? Remote sensingdata allows us to investigate a largearea in a timely and cost-efficientmanner. Features indicative of recentfaulting include fault scarps developedon young surfaces or displacing younggeological deposits, sag ponds and/orspring lines, displaced drainages, andother ‘anomalous’ geomorphicfeatures. Digital data and imageprocessing tools allow us to manipulatetopographic data, in particular alteringthe sun angle on Digital ElevationModels (DEMs), to enhance subtlemorphotectonic features and betterlocate surface rupturing faults.Coupled with detailed seismicityrelocation studies, this becomes apowerful tool in locating and identifyingactive faults.

Once potentially active faults havebeen identified, aerial and fieldreconnaissance investigations arecommonly the next step in seismicsource characterisation. As well asinves t iga t ing fau l t -spec i f i cgeomorphology, regional landscapefeatures, including mountain rangefronts and drainage basin geometrycan give an insight into the grosscharacteristics of recent tectonicmovements, e.g., sense ofdisplacement, strain partitioning (ifany), total offset, and long-term sliprates. This also allows the selectionof localities for site-specificinvestigation. By carrying out detailedsite-specific investigations, mostcommonly palaeoseismic trenching,we can build up a picture of thebehaviour of different faults, includingtheir slip rates and/or interseismicrecurrence intervals, style andmagnitude of displacement, and byway of fault rupture length, a measureof the maximum earthquake possibleon any given seismic source. Carefulselection of trenching localities,detailed stratigraphic logging and

reliable, accurate age-dating are thekey factors in successfulpalaeoseismic investigations where weaim to determine fault recurrence and/or slip rate. Without age-dating all weare left with is, at best, an estimate ofslip-per-event, and no temporal contextwithin which to place these data. Onits own, slip-per-event data can beused to determine earthquakemagnitude, but it does not answer ourthird seismic source characterisationquestion: How often?

It is important to understand and,where possible, quantify theuncertainty in the measurement ofeach of these source characteristics.An ideal trenching site will have had along history of faulting coupled withcontinuous sedimentation to record acontinuous history of offsets. In anideal World, the most recent faultingevent (surface rupturing fault scarp) willbe draped with a dateable horizonallowing an estimation of the time sincethe last event; in addition to a robustpalaeoseismic chronology, these datacan be used to enhance our estimatesof probabilistic hazard assessment toinclude a time-dependent factor. Theinterpretation of multiple trenchexposures along any given fault allowsus to build up a model of long-term faultbehaviour, including developing anunderstanding of fault segmentationand recurrence. Only by suchextensive and scientifically rigorousfield investigations can we begin totackle some of the more difficultquestions in seismic sourcecharacterisation, such as temporalscaling; how do we resolve differencesbetween geodetically measured strain,instrumental and historical seismicitycatalogues, and paleoseismologicalfault slip and/or recurrence intervals?Are these differences real, indicatingtime-variant fault behaviour, or are theymerely a reflection of the relativecompleteness of each recordingperiod? PSHA allows us to entertainmultiple seismic source modelsincorporating a multitude of faultbehaviour.

Detailed seismotectonic andpalaeoseismic investigations allow usto obtain the data for seismic sourcecharacterisation. Fully understandingthe limitations and uncertainties inthese data allows the development of

a comprehensive and defendableseismic source model, thereby leadingto a realistic understanding of seismichazard at a site. Thoroughinvestigation of the seismotectoniccharacter of a region, leading to acomprehensive suite of seismic sourcemodels, fully describing theiruncertainties, should be the goal for allPSHAs. Coupled with rigoroussensitivity analysis, this should lead tomore robust, believable hazardevaluations.

DiscussionThe discussion which followed thispresentation began with Bryan Skippraising the issue of the conversesituation where effort is directed attrying to prove that a fault is NOT activeunder the current tectonic regime. Theneed to search for cross-cutting andhierarchical relationships and fordateable materials was mentioned anda warning given concerning the ‘blackarts’ involved in some datingtechniques. In particular, it should beensured that the technique beingemployed is such that the ‘clock’ it useswill necessarily have been reset by themost recent fault movement since thisis not always the case.

Ian Smith wondered about thequantitative merits of the output frompalaeoseismic studies and was toldthat the geometrical dimensions of thefault could be used to assess quitereliable magnitudes but that recurrenceintervals were more difficult toestimate.

The applicability of palaeoseismicmethods in offshore environments wasraised by Phillip Cooper and it wasagreed that, in such situations, reliancehad to be placed in remote geophysicalmethods, although the precision ofsuch methods was always improving!Bryan Skipp recalled some of theproblems experienced by geophysicalcontractors in dealing with engineerswhose specified requirements borelittle relation to the output that couldbe achieved in practice or to theuncertainties associated with suchtechniques. David Mallard quoted aninstance where fairly minorperturbations in the overlyingdepositional characteristics producedseismic sections that might have been


interpreted as implying very recentmovements on an old fault.

Robin Adams drew attention to thelimitations of palaeoseismic methodsin environments where intermediatedepth events (ca. 100km) provide thegreatest contribution to seismic risk (forexample, mega events at 30 to 50kmon subduction zones). Clark Fentonreplied that the paleoseismology ofsubduction zones could be unravelledand referenced the work of JohnAdams (Adams, 1990) and ChrisGoldfinger (Goldfinger et al., 2003) inusing detailed age-dating of deepocean turbidites that occur overgeographically wide areas to indicatelarge subduction zone events. Thework of Kerry Sieh and his students atCaltech on coral growth rings(Zacharien et al., 1999) indicatingsubsidence/uplift histories of reefsalong subduction zones was also veryuseful in identifying both historical andpaleoseismic events.

Questioning Clark Fenton’s whole-hearted support for the probabilisticapproach to hazard assessment, DavidSmith noted that palaeoseismicevidence might be seen as fitting morecomfortably into deterministic methodswhere periodicity is of no concern. Inthe same connection, Julian Bommerpointed out that the origins of thedeterministic method sprang preciselyfrom the desire to incorporategeological evidence into hazardstudies. He accepted that the mostrational approach to the uncertaintiesinvolved in hazard assessment is touse logic-tree methods but consideredthat there is no reason why a logic-treeapproach should not be used to derivedeterministic hazard estimates.Notwithstanding these comments,Clark Fenton continued to advocateprobabilistic methods with thestipulation that it was essential to carryout a disaggregation of the results toinvestigate the contribution of individualseismic sources.

Jon Hancock raised the issue ofsafety aspects in palaeoseismicinvestigations and was reassured thatthe statutory safety precautions shouldalways be employed, for example,when trenching. Clark Fentonmentioned examples of hydraulically-shored and propped trenches of the

order of 4.5m depth on the MaacamaFault and stepped and shored trenchesup to 20m for the Wasatch Fault.

The question as to whether any newhi-tech advanced techniques werecoming into the field of palaeoseismicinvestigations was raised by Ian Smithwho was told that LIDAR wasbecoming increasingly useful in faultmapping, especially for small tectonicgeomorphological features and inregions with dense vegetation cover.The recent work of Carol Prentice (USGeological Survey) and co-workerswas cited as an example. GroundPenetrating Radar was also cited as apotentially useful technique, althoughthe limitations of this method,particularly in water-saturated clayeysoils, were emphasised.

Antonio Pomonis raised a query as tothe geological timescale over whichpalaeoseismic techniques might beexpected to provide useful evidence:could they be used, for example, invery old rocks? Clark Fentonanswered that as long as there was agood, continuous, geological record,the paleoseismic record should, intheory, be limitless! However in reality,erosion, non-deposition and thelimitations of high-precision age-datingtechniques limited ‘good’ paleoseismicdata to about 100,000 years. Whereproblems are most likely to arise is withfaults that have very long recurrenceintervals. The Ungava, Canada,earthquake (Adams et al., 1991) wasquoted as an extreme example, wherea Proterozoic ductile shear zone, withno history of any Phanerozoicmovement generated a surface-rupturing earthquake in 1989!

Ed Russell raised the issue of strainmonitoring and was told that suchmeasurements could be useful inhighly active areas. Problems canoccur, however, where GPSinstallations are monitored for shortperiods since such observations canappear to be inconsistent with thedeformation rates indicated by other,longer term, diagnostics.

Tiziana Rosetto asked for confirmationconcerning the minimum sizes ofearthquakes that could be expected toleave geological evidence. In ClarkFenton’s opinion, events of Moment

Magnitude (M) 6¼ would be areasonable threshold for an environmentlike the Basin and Range region in theUSA, whereas the figure might be M 6½in an environment like the Californiancoastal belt. For an overall figure,covering almost all circumstances, hewould suggest 6½ ± ¼.

Following this pronouncement, themeeting closed.

References

Adams, J., 1990, Paleoseismicity ofthe Cascadia subduction zone:evidence from turbidites off theOregon-Washington margin:Tectonics, v. 9, p. 569-583.

Adams, J., Wetmiller, R.J., Hasegawa,H., and Drysdale, J., 1991, The firstsurface faulting from a historicalintraplate earthquake in North America:Nature, v. 352, p. 617-619.

Goldfinger, C., Nelson, C.H., andJohnson, J.E., 2003, Holoceneearthquake records from the Cascadiasubduction zone and northern SanAndreas fault based on precise datingof offshore turbidites: Annual Reviewof Earth and Planetary Sciences, v. 31,p. 555-577.

Zachariasen, J., Sieh, K., Taylor,F.W., Edwards, R.L., and Hantoro,W.S., 1999, Submergence and upliftassociated with the giant 1833Sumatran subduction earthquake:Evidence from coral microatolls:Journal of Geophysical Research,v. 104, p. 895-919.


The Estimation of Peak Ground Acceleration in the UK after an Earthquake

IntroductionFor more than thirty years, the BritishGeological Survey (BGS) has beenundertaking instrumental seismicmonitoring of the UK, as a result ofwhich a considerable database ofrecords has been collected. Most of thedata consists of very low level readingstaken from instruments that measurevelocity. A smaller set of data consistsof higher level readings that are takenfrom instruments that record peakground acceleration (pga). It is this setof data and its treatment that is thesubject of this article. The articledescribes the regression of the dataand concludes with the discussion ofthe application of the regression resultsto the estimation of pga at UK siteswhere this may be of use. Such sitesare likely to be those where a seismicsafety case has been required.

Throughout this article magnitude isexpressed in terms of Local Magnitude,distance in kilometres and peak ground

acceleration in %g. Furthermore, noattempt has been made in the reportedwork to explore the site conditions ateach of the measurement stations.This is clearly a possibility for thefuture.

Treatment of DataThe peak ground acceleration datacollected to date by BGS is listed inTable 1. This data has been subject toa straightforward multi-variable linearregression using the relationship

Ln (pga) = A.M + B.Ln(R) + C…………………………………… (1)

Where:• pga = the peak ground acceleration

in units of %g.• M = the Local Magnitude of the

Earthquake• R = the focal distance in units of km.

A, B and C are the constantsdetermined by the regression analysisof the data. The pga data provided isthe larger of the two measuredhorizontal orthogonal components.

The regression analysis wasundertaken using a proprietarystatistical package (Axum) and wascross- checked using two separatemethods in the computer programMathCad. The constants obtained fromthe regression analysis give rise to thefollowing relationship.

Ln (pga) =1.879 . M - 1.254.Ln(R) - 3.969............................……………… (2)

The standard error of the regressionis 1.054. The computation of this valuehas also been checked using anindependent method in Mathcad.Figure 1 provides a three dimensional

A straightforward method, based on UK instrumental data, is offered by Chris Allen

Figure 1 Three Dimensional Data Plot


plot of the data and the regressionplane. Figure 2 provides the same plot,except that the view of the plot is alongthe regression plane.

Application of the Above WorkThere are many sites within the UK thathave undertaken a formal seismicassessment and therefore have inplace a safety case for earthquakeloading. There may be interest at suchsites in the levels of pga experiencedafter an earthquake, either as a resultof internal, or external, queries. Thelatter may become more frequent as aresult of the implementation of theFreedom of Information Act.

Such sites may not have seismicmonitoring equipment installed, or, ifseismic monitoring equipment is inplace, it may be such as to have adefined trigger level, below which noreading is recorded.

It is considered that the attenuationexpression described herein and givenin equation 2 above may be of use forsuch sites. It is a straightforward matterto incorporate equation 2 in aspreadsheet. This can also contain theOrdnance Survey coordinates of thesites of interest. After a significant UKearthquake, BGS will quickly provide

the Local Magnitude, the focal depthand the geographical location of theepicentre. The spreadsheet cancalculate the focal distance to the sitefrom the geometry as shown in Figure3 and use equation 2 to determine themean pga at each site of interest and,if required, the pga corresponding tothe mean plus one standard deviationlevel.

Each site can then use this informationto provide reassurance of the levels ofpga experienced and to compare with

that which forms the basis of the formalseismic assessment of the site.

However, in any application of this workalong the lines suggested, there wouldremain the matter of the range ofmagnitude and epicentral distancedata for which the attenuation equationcould be deemed to be valid. This topicis left to be resolved at the time of anyfuture application of this work. Clearlymagnitudes of above 4.7 andepicentral distances below a few 10’sof kilometres may be outsidereasonable limits of this attenuation

Figure 2: In Regression Plane View

Figure 3: Basis for Calculation of Focal Distance


equation. On the other hand, the limitof applicability may to some extent beself- correcting, as the occurrence of,say, a magnitude 5 at some time in thefuture may produce hard data that willallow extension of the limits ofapplicability.

The FutureIt is anticipated that the UK seismicmonitoring project run by BGS willcontinue. The number of seismicinstruments that can record moderatelevels of peak ground acceleration hasincreased significantly over theduration of the project, with the resultthat the amount of acceleration datashould increase with time more quicklythan has happened in the past. It wouldthus be sensible to modify theregression analysis as discussedherein when new data arrives. This isan activity that the author may pursue.It would be sensible if others followedsuit.

It is recognised that currently availabledata has been analysed on a basis thatis relatively straightforward. There maybe other approaches that are moreelegant and perhaps better suited tothe task in hand. The work herein isthus offered on the basis that theremay be merit in the promulgation of thework discussed herein. Futurealternative treatments of this topic arethus not excluded; indeed they areencouraged.

Information on the UK seismicmonitoring project may be obtained bycontacting Dr. Brian Baptie at the BGSEdinburgh office at Murchison House.Contact may be made via the BGS website, www.bgs.ac.uk/contacts.

The permission to use the datacollected by BGS is gratefullyacknowledged.

Date Earthquake Local Focal PGALocality Magnitude Distance (%g)

(km)19940317 Newtown 3.1 64 0.04319940512 Stratford 3.0 91 0.02419940611 Constantine 2.2 10 0.14719940817 Isle Of Skye 3.1 18 0.04019960307 Shrewsbury 3.4 86 0.04519960920 Llandrindod 3.0 38 0.05519961110 Penzance 3.8 35 0.53719970622 Jersey 2.2 21 0.09819971108 L Maree 2.5 27 0.04219980403 Annan 1.1 7 0.09819980528 Wigton 1.5 21 0.02419980721 Locharbriggs 2.0 31 0.00419990121 Boston 2.8 96 0.10319990304 Arran 4.0 137 0.04119990314 Dumfries 1.9 30 0.00319990617 Hereford 2.8 39 0.20419990713 Jersey 1.8 22 0.06919990903 Johnstonebridge 2.1 26 0.00819991025 Sennybridge 3.6 40 0.38520000107 Dumfries 1.8 30 0.00320000212 Loghilphead 2.7 166 0.00220000424 Calthwaite 2.6 41 0.07220000622 Lleyn Peninsula 2.6 53 0.03320000808 Middlesbrough 2.7 127 0.00320000923 Warwick 4.2 77 0.21220000923 Warwick 4.2 88 0.06720000923 Warwick 4.2 102 0.06820010513 Dumfries 2.9 29 0.04820010531 Off Hartland Pt, Devon 3.6 102 0.07620010627 Sedbergh 2.2 103 0.002200110 Bargoed 3.0 42 0.16320011028 Melton Mowbury 4.1 19 2.16120011028 Melton Mowbury 4.1 144 0.14520020922 Dudley 4.7 82 0.21420020922 Dudley 4.7 84 1.44820020922 Dudley 4.7 285 0.10220020922 Dudley 4.7 117 0.19420020922 Dudley 4.7 230 0.15320020922 Dudley 4.7 148 0.18320021021 Manchester 3.9 100 0.06120021021 Manchester 3.5 101 0.06320021029 Annan 1.8 19 0.00820030819 Doncaster 3.1 68 0.05820040229 Oldham 3.1 42 0.06420040229 Oldham 3.1 99 0.06220040229 Oldham 3.1 180 0.00120040629 Lockerbie 2.0 180 0.01720040807 Dumfries 2.3 29 0.00920041013 Eskdalemuir 1.7 21 0.00520041027 Eskdalemuir 1.3 22 0.00220041103 Eskdalemuir 2.7 22 0.03320041103 Eskdalemuir 1.8 21 0.00520041103 Eskdalemuir 2.1 21 0.00820041103 Eskdalemuir 1.3 21 0.00220041104 Eskdalemuir 1.2 21 0.00120041105 Eskdalemuir 1.9 21 0.00620041128 Eskdalemuir 2.9 23 0.023

Table 1: UK Earthquake Acceleration Data

BGS Strong Motion Data


Blast Intensity

At 06h01m on 11 December 2005, ahuge explosion occurred at theBuncefield fuel depot near HemelHempstead, causing much damage,and starting a fire that required severaldays to bring under control. Theexplosion is considered to have beenthe largest in post-war UK history. Theblast was clearly detected by the UKseismic monitoring network atdistances of up to 300 km, and theseismic record shows well the passageof both the ground-transmitted shockwave and the air wave.

The engineering impact of the blast onstructures in and near the depot wasthe subject of an immediateinvestigation, the results of which at thetime of writing are understood not yetto be available. However, because ofthe newsworthiness of the event, therewas considerable public and mediainterest in the far-field effects of theblast. Such things are not normallystudied, and no procedures are inplace for handling such aninvestigation. In order to gather dataquickly, and not miss this (thankfullyrare) opportunity, the British GeologicalSurvey (BGS) took matters in hand byimplementing a survey of the widereffects of the blast using theprocedures set up for handlingmacroseismic investigations ofearthquakes. This involved setting upthe BGS web questionnaire page forthe Buncefield event; the relevant pagewas linked from the BBC’s news storyon the BBC web pages. About 2000responses were received within 24hours, and a total of 3016 responseswere received before the page wasclosed at the end of the month.

While this was a good result, themanner of collecting the data was notideal. There was no time to composea web-enabled questionnairespecifically directed towards surveyingblast effects, which is a difficult task inthe absence of any standard model ofblast questionnaire. Although there aremany commonalities between theeffects of explosions and earthquakes,

such that the one is often mistaken forthe other (at least, at far distances),there are also significant differences,reflecting the fact that in one casetransmission is primarily through rock,and in the other, primarily through air.

This also affects the processing of thedata. It is quite possible to take themass of questionnaire data gatheredfor the Buncefield explosion andprocess it according to standardmacroseismic procedures, as thoughit were earthquake data, and arrive atintensity values on the EuropeanMacroseismic Scale (EMS – seeGrünthal 1998). Indeed, this was done.However, the meaning of the values isquestionable, in that one is using a toolfor measuring earthquake effects andapplying it to blast effects.

The results are shown in Figure 1. Thismap covers only the south-east ofEngland; a few isolated reports camein from as far away as Somerset,Wales and Yorkshire, but the credibilityof some of these outliers is doubtful.The intensity values do show a

coherent pattern, and are as regularas any earthquake data set in terms ofattenuation with distance. One can alsonote a directivity in the blast wherebythe strongest effects are biasedtowards the south-south-west. In termsof what the EMS intensities really meanfor an explosion, the interpretation inTable 1 can be proposed.

As far as I am able to determine, thereis no published intensity scaledesigned for dealing with blast effects.This seems a surprising omission.Presumably, in the past, thoseinvestigating blast effects within thedamage zone have been content toenumerate the damaged buildings, andin the far field, either no-one has beeninterested in surveying the distributionof effects, or if they have, anearthquake intensity scale has beenpressed into service as in the case ofBuncefield. However, the possibility ofproducing blast effect maps isinteresting, and in these days whereterrorist explosions are a constantthreat, possibly useful as well.

Roger Musson reports on how procedures normally used for investigating macroseismic eventshave been used to investigate the December 2005 Buncefield explosion. A Blast Intensity Scale is

proposed on which further contributions would be welcomed.

Figure 1 EMS “intensity” map of the Buncefield explosion.


The following scale is tentativelyproposed as a working draft for aproper blast intensity scale. It is basedon a mixture of the work of Glasstoneand Dolan (1977) and personalexperience; it undoubtedly needsrefinement and improvement fromengineers with first-hand professionalexperience in dealing with the effectsof large explosions. My aim inpublishing it here is to attract attentionin the hope that contributions will beforthcoming in this regard. It is notpossible to use it with the data so farcollected for Buncefield, partly becausethe data would need to be collected ina way that envisaged the scale to beused for the analysis, and alsobecause obviously the higher degreescan only be assessed from fieldinvestigations.

The qualities that make any intensityscale good or bad are discussed atlength in Musson and Grünthal (2006).In deciding the best number of intensitydegrees to have in a scale, and howthey should be defined, considerationneeds to be given to the followingissues:

· Discrimination: can the distinctionsbetween different degrees of thescale be clearly recognised in thefield?

· Consistency: do all the diagnosticsfor any degree of the scale reallybelong together at the same levelof effect?

· Regularity: does the decrease inintensity with each lower degree inthe scale follow an orderlyprogression of decreasing effects?

It is in the interests of discriminationthat the number of degrees has beenlimited to seven, but it might be felt that,for instance, a degree could beinserted between 4 and 5. It requiresexperience with actual data todetermine whether the scale is moreregular with or without such a degree.As with earthquake intensity, when adata set is mapped, a good scaleshould show even logarithmic spacingof contours. If one consistently findsthat some contours are bunched inviolation of this, there is a fault in thescale construction. For the higherdegrees of the scale I have beenentirely reliant on Glasstone and Dolan(1977) as to the consistency of theproposed definitions.

Comments from the blast engineeringcommunity are invited. Whether or notsuch a scale has scientific applicationsin the investigation of explosions, itshould certainly be useful in terms ofcommunicating results to the public,

who, in the case of Buncefield event,were very interested to know over whatarea the blast extended.

R.M.W. MussonBritish Geological Survey, West MainsRoad, Edinburgh, EH9 3LA, UK

AcknowledgementsThis work was supported by theNatural Environment ResearchCouncil (NERC). The UK seismicmonitoring programme is supported bya Customer Group led by the ODPM.This paper is published with thepermission of the Executive Directorof the BGS (NERC).

ReferencesGLASSTONE, S, and DOLAN, P J.1977. The effects of nuclear weapons.(Washington D.C.: US Department ofDefense and Department of Energy.)

GRÜNTHAL , G (editor). 1998.European Macroseismic Scale 1998.Cahiers du Centre Europèen deGèodynamique et de Seismologie. No.15. (Luxembourg: Conseil del’Europe.)

Musson, RMW and GRÜNTHAL , G.2006. On the comparison of intensityscales, in preparation.

EMS value Significance

6 Strong blast, some damage caused5 Blast sufficiently strong to be frightening or alarming4 Observed by many people in the locality2 or 3 Observed by a few people

Table 1 - Suggested interpretation of EMS intensity values for blasts

Blast Intensity Scale 2005 (Provisional Draft)1. No detectable effect

2. A distant sound is heard; sometimes, depending on atmospheric conditions,shaking of houses will be perceptible, even without sound.

3. A loud sound is heard; buildings may shake; windows and slates may rattle.

4. Vibration of the ground is apparent even outdoors. Many windows areshattered.

5. Masonry buildings suffer cracking to walls; interior partitions may collapse;wood frame buildings collapse partly or wholly; roofs are damaged.

6. Masonry buildings suffer collapse; truss bridges and similar constructionsare distorted and may collapse in some cases.

7. Modern steel frame buildings are badly damaged and may collapse; girderbridges collapse; cars and lorries are overturned or hurled in the air.

The next biennial Mallet Milnelecture will be given by ProfessorRobin Spence and will take place on30th May 2007 at the Institution ofCivil Engineers. The lecture will beentitled “Saving lives inearthquakes: successes andfailures of seismic protection since1960”.

Mallet Milne Lectureand Reception 2007


Seismic Design to Eurocode 8

SECED-Imperial College Short Course with Design Workshops21-22 September 2006 at Imperial College London

SECED will be organising a short course jointly with Imperial College Londonon 21 and 22 September 2006. The two-day course will take the form of aseries of morning lectures from leading experts from both the university andconsulting sectors, followed by practical design workshops on Eurocode 8 inthe afternoon. The course will link with initiatives established by the Institutionsof Civil and Structural Engineers for assisting with the implementation of theStructural Eurocodes. The programme and details are now available. Forfurther information contact Dr Ahmed Elghazouli at Imperial College London(e-mail: [email protected]).

Last year’s SECED Young Engineers Conference, held at the University of Bath, proved to be a truly internationalevent. Whilst the focus was on UK research and practice, there were attendees from across Europe, Iran and as far afield as Hong Kong and Malaysia. The aim of the conference was to provide a link between young researchers andpractitioners working in the fields of earthquake engineering, seismology and civil engineering dynamics. The livelydiscussions and debate which followed each presentation were an indication of the success of this aim. The presentationscovered a wide range of topics, but the common themes which emerged were soil liquefaction, advanced structuraland geotechnical test methods, vulnerability assessment of low engineered structures, the development and assessmentof realistic ground motion models and methods of mitigating structural vibrations. Additionally there was a small amountof work presented on human-structure interaction and wind-structure interaction.

Of particular poignancy, with the Conference coming soon after the tragedy of the Asian tsunami, was a presentation byYusoff Nor Azizi. Having come from Malaysia, Yusoff witnessed the devastation of the tsunami first hand and precededhis planned presentation with a discussion of the effects of the tsunami on the Malaysian population and infrastructure.This was a theme later picked up on by SECED Chairman, Zygmunt Lubkowski, in his speech following the ConferenceDinner, laying down the challenge to us, as young engineers, to help prevent similar tragedies from occurring.

We were especially honoured to have two key-note speakers at the conference. From the academic side, GopalMadabushi from Cambridge University, arriving straight from a trip to America, spoke about the latest advances inphysical and numerical geotechinical modelling in the context of earthquakes. The talk focussed upon developmentsin centrifuge testing as a tool for validating numerical models. In particular the provision of realistic boundary conditionsand evaluating strain fields and soil deformation for such physical tests were discussed. From the industry side ofearthquake engineering, Edmund Booth gave an entertaining and unique insight into the drafting of the Europeanseismic code, Eurocode 8. As a member of the drafting committee, he was well placed to describe the complexEuropean politics and red-tape which surround the development of such a code. He also discussed the basis on whichthe code was developed and the impact of codes and standards on as-built seismic safety.

Prizes were given for the best papers of the Conference. Highly Commended awards of £50 were presented to JackieSim, from the University of Oxford, for her paper entitled “Response of a Joint Passive Crowd-SDOF System Subjectedto Crowd Jumping Load” and to Paul Murtagh, from Trinity College Dublin, for his paper entitled “Effect of AerodynamicDamping on the Damped Natural Frequencies and Mode Shapes of Towers Supporting Utilities”. However, the £100prize for the best paper of the conference went to Nick Sartain and Kubilây Hiçyilmaz, of Ove Arups and Partners, fortheir paper entitled “Key Issues In the Seismic Design of Immersed Tube Tunnels”. Their paper really captured theessence of the Conference, combining elements of seismology, geotechnics and structural dynamics, applied to aninteresting practical problem where a rigorous but pragmatic investigation is critical. The paper discussed the problemsinvolved in such a situation and the benefits of various available analysis methods and their role in risk assessment.The paper seemed to summarise, in an applied sense, the recurring themes that were focussed on by other practitionersand researchers throughout the conference.

The Conference organisers wish to thank all those who contributed to the success of the conference. A limited numberof copies of the conference proceedings are available on CD-ROM from SECED at a cost of £15.00. Please contactthe SECED secretary for more information.

Antony Darby, University of Bath

Report on the SECED Young Engineers Conference, March 2005

29 March 2006Force and Displacement Based VulnerabilityAssessment for Traditional BuildingsICE 6.00pm

26 April 2006AGM and This Year’s Earthquake:Pakistan Earthquake - October 8th 2005ICE 5.30pm

Forthcoming Events


SECED NewsletterThe SECED Newsletter is publishedquarterly. Contributions are welcome andmanuscripts should be sent on a PCcompatible disk or directly by Email.Diagrams, pictures and text should be inseparate electronic files.

Copy typed on paper is also acceptable.Diagrams should be sharply defined andprepared in a form suitable for directreproduction. Photographs should behigh quality (black and white prints arepreferred). Diagrams and photographsare only returned to the authors onrequest.

Articles should be sent to:

John Sawyer,Editor SECED Newsletter,c/o The Secretary,SECED,Institution of Civil Engineers,Great George Street,LondonSW1P 3AA, UK.

E: [email protected]

SECEDSECED, The Society for Earthquake andCivil Engineering Dynamics, is the UKnational section of the International andEuropean Associations for EarthquakeEngineering and is an affiliated society ofthe Institution of Civil Engineers.

It is also sponsored by the Institution ofMechanical Engineers, the Institution ofStructural Engineers, and the GeologicalSociety. The Society is also closelyassociated with the UK EarthquakeEngineering Field Investigation Team. Theobjective of the Society is to promote co-operation in the advancement ofknowledge in the fields of earthquakeengineering and civil engineeringdynamics including blast, impact and othervibration problems.

For further information about SECEDcontact:The Secretary,SECED,Institution of Civil Engineers,Great George Street,London SW1P 3AA, UK.

SECED WebsiteVisit the SECED website which can befound at http://www.seced.org.uk foradditional information and links to itemsthat will be of interest to SECEDmembers.Email: [email protected]

NOTABLE EARTHQUAKES OCTOBER – DECEMBER 2005Reported by British Geological Survey

YEAR DAY MON TIME LAT LON DEP MAGNITUDES LOCATIONUTC KM ML MB MW

2005 1 OCT 22:19 16.64S 70.79W 20 5.3 SOUTHERN PERU

2005 8 OCT 03:50 34.54N 73.59E 26 6.9 7.6 PAKISTANAt least 86,000 people killed, more than 69,000 injured and extensive damage in northern Pakistan. Theheaviest damage occurred in the Muzaffarabad region of Kashmir where entire villages were destroyed andUri where 80% of the town was destroyed. Around 4 million people in the region were left homeless asresult of this earthquake.

2005 15 OCT 04:24 34.01N 74.00E 10 5.2 PAKISTANTwo people killed in the Uri area.

2005 15 OCT 15:51 25.32N 123.36E 183 6.2 6.5 TAIWAN REGION

2005 16 OCT 07:05 36.01N 139.77E 40 5.1 HONSHU, JAPANTwo people injured in the Tokyo area.

2005 20 OCT 21:40 38.15N 26.75E 10 5.5 5.9 WESTERN TURKEYOne person killed and 15 injured at Izmir. Minor damage to several buildings at Urla.

2005 27 OCT 11:18 23.60N 107.80E 10 4.2 GUANGXI, CHINAOne person killed, another injured at Bose and several homes damaged at Taiping.

2005 29 OCT 04:05 45.21S 96.90E 8 6.1 6.5 SE INDIAN RIDGE

2005 6 NOV 02:11 34.52N 73.39E 10 5.2 PAKISTANSeven people injured at Batgram.

2005 8 NOV 07:54 9.97N 108.29E 10 5.1 SOUTH CHINA SEAOne person killed at Ho Chi Min City, Vietnam.

2005 8 NOV 21:06 55.22N 3.11W 3 0.6 ESKDALEMUIRFelt in the Langholm, Dumfries and Galloway region (2-3 EMS).

2005 14 NOV 21:38 38.11N 144.90E 11 6.7 7 HONSHU, JAPANA tsunami wave with a maximum wave height of 32cm was recorded at Ofunato.

2005 17 NOV 19:26 22.32S 67.89W 163 6 6.9 BOLIVIAPower outages occurred throughout Tocopilla, Chile.

2005 19 NOV 14:10 2.16N 96.79E 21 6 6.5 SIMEULUE, INDONESIA

2005 26 NOV 00:49 29.70S 115.69E 10 5.4 5.2 JIANGXI, CHINAAt least sixteen people killed, 8,000 injured and over 150,000 houses destroyed in the Jiujiang andRuichang areas. Minor damage also occurred in Wuhan.

2005 27 NOV 10:22 26.77N 55.86E 10 6.1 6 SOUTHERN IRANThirteen people killed, around 100 more injured and seven villages severely damaged on Qeshm and morethan 80% of the buildings in Zirang destroyed.

2005 2 DEC 13:13 38.09N 142.12E 29 6.1 6.5 HONSHU, JAPAN

2005 5 DEC 12:19 6.23S 29.78E 22 6.3 6.8 LAKE TANGANYIKAAt least six people killed, several others injured and more than 300 houses and a church destroyed atKalamie, Congo.

2005 10 DEC 23:21 56.84N 5.22W 8 3 FORT WILLIAMFelt throughout the Fort William, Highland region. The felt area for this event is approximately 7,300 km2(2-3 EMS), 1,100 km2 (4 EMS) and 215km2 (5 EMS).

2005 11 DEC 14:20 6.58S 152.20E 10 6.1 6.6 PAPUA NEW GUINEA

2005 11 DEC 23:14 53.11N 0.02W 3 2 BOSTON, LINCS

2005 12 DEC 21:47 36.36N 71.11E 224 5.9 6.6 AFGHANISTANFive people killed in Tili, one person injured in Jalalabad and around 100 houses damaged and 300 livestockkilled in Badakhshan.

2005 13 DEC 03:16 15.26S 178.57W 10 6.1 6.7 FIJI ISLANDS REGION

2005 14 DEC 03:30 53.00N 5.64W 10 2.8 IRISH SEAFelt coast of Wicklow, Ireland (3 EMS).

2005 23 DEC 03:25 56.68N 5.68W 7 2.7 LOCH SUNARTFelt Strontian, Highland region (3 EMS).

2005 23 DEC 04:58 56.67N 5.66W 8 2.4 LOCH SUNARTFelt Strontian, Highland region (3 EMS).

2005 24 DEC 02:01 35.23N 136.84E 43 4.5 HONSHU, JAPANOne person injured in Yokkaichi.

2005 29 DEC 04:40 56.22N 3.76W 8 1.2 GLENDEVONFelt Glendevon, Central region (3 EMS).

2005 31 DEC 22:40 56.27N 3.77W 6 2.5 BLACKFORDFelt Blackford and Auchterarder, Tayside region (4 EMS).

Issued by: Davie Galloway, British Geological Survey, February 2006.Non British Earthquake Data supplied by: The United States Geological Survey

Documents

S THE SOCIETY FOR EARTHQUAKE AND CIVIL E S E C E D ... · THE SOCIETY FOR EARTHQUAKE AND CIVIL ENGINEERING DYNAMICS NEWSLETTER Volume 19 No 2 March 2006 ... Page 10 Mallet Milne Lecture