The dynamic behavior of shallow marine reservoirs Insights from the Pliocene of offshore North

The dynamic behavior of shallow marine reservoirs: Insights from the Pliocene of offshore North TrinidadNigel E.Cross,Zana K.Williams,Arman Jamankulov, Candice E.Bostic,Valini C.Gayadeen,Helisaul J.Torrealba,and Elizabeth S.DraytonA B S T RA C TThe Pliocene reservoirs of the North Coast Marine Area(block NCMA-1)comprise shallow marine shoreface to shelf sandstones (up to30m[98ft]thick)and interbedded shelfal mudstones organized into stacked upward-coarsening parasequences (ca.25–70m½∼82–230ft thick).Production behavior is strongly conditioned by the interplay of subregional sand distribution and aquifer connectivity,and intrafield reservoir architecture com-plexity,commonly in the form of clinoforms.The M2parasequence is a well-connected reservoir unit and forms part of a late highstand systems tract with a strongly progra-dational rge-scale reservoir connectivity can be demonstrated by pressure communication in producing fields up to10km(6mi)apart.The widespread nature of the sand also links the reservoir to a significant aquifer that provides pressure support during field depletion.In contrast,clinoforms that domi-nate the M2reservoir architecture at a variety of scales reduce lat-eral connectivity.Wells situated in individual sand bodies show significant pressure interference during production,whereas wells that straddle bounding clinoforms(possible parasequence boun-daries),although in pressure communication,show more subtle pressure interference.In the Chaconia field,clinoforms that bound lobate sand bodies in the M2reservoir have also apparently slowed the onset of water cut within the field by several years and therefore increased reserves recovery.Similarly,in the M6 reservoir of the Ixora field a significant bounding clinoform has partitioned the field through much of its production life,but pro-duction data are now suggesting breakdown and late-stage Copyright©2015.The American Association of Petroleum Geologists.All rights reserved. Manuscript received September6,2013;provisional acceptance November25,2013;revised manuscript received April14,2014;final acceptance September18,2014.DOI:10.1306/09181413165A U T H O R SNigel E.Cross∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;present address: Mubadala Petroleum,Level27,Al Maqam Tower, Sowwah Square,Al Maryah Island,P.O.Box4887, Abu Dhabi,UAE;nigel.cross@ Nigel Cross has a B.Sc.degree and a Ph.D.from Royal Holloway,University of London.He has worked for BG Group since2004first in Egypt and later in Trinidad and Tobago.Prior to BG,he worked for Petro-Canada,Hess,and Badley Ashton. He is currently the NCMA subsurface team leader, and his technical interests include rift basin geology and all aspects of integrated reservoir modeling. Zana K.Williams∼BG Group,100Thames Valley Park Drive,Reading,Berkshire,RG61PT, United Kingdom;present address:QGC,275 George Street,Brisbane,Queensland,4000, Australia;Zana.Williams@Zana Williams has a B.Sc.degree from the University of Leeds and a Ph.D.from The University of Edinburgh.She has worked for BG Group since 2007first as a new ventures geologist in the United Kingdom and lately as a development geologist in Trinidad and Tobago and Australia.Arman Jamankulov∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;Arman. Jamankulov@Arman Jamankulov has a B.Sc.degree in reservoir engineering from Kazakh National Technical University,Kazakhstan,and has a M.Sc.degree in petroleum engineering from Delft University of Technology,The Netherlands.He has worked for BG Group since2007first in Kazakhstan,United Kingdom(UK)and later in Trinidad and Tobago.Prior to BG,he worked for Kazakh National Oil and Gas Company. Candice E.Bostic∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;Candice.Bostic@ Candice Bostic has a B.Sc.degree from the University of the West Indies,Trinidad.She has worked for BG Trinidad and Tobago since2007 as a geophysicist in new ventures and currently in the NCMA subsurface team.Valini C.Gayadeen∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;Valini.Gayadeen@ AAPG Bulletin,v.99,no.3(March2015),pp.555–583555pressure support across the same clinoform.The M6parasequen-ces form part of a lower quality early highstand systems tract and are more discontinuous than their M2counterpart due to well-developed bounding clinoforms.As such they are not char-acterized by large active aquifers and do not benefit from such prominent long-term pressure support.The M4reservoir is an iso-lated sand-body complex that forms part of a high-frequency low-stand complex.It shows similar stratigraphic heterogeneity to that displayed in the M2,with variations in pressure decline exhibited within individual wells of the same reservoir unit.Its isolated geometry means it is not connected to a regional aquifer and is therefore undergoing volumetric decline with no water breakthrough.INTRODUCTIONShallow marine clastic reservoirs are an important component to a number of the world ’s largest hydrocarbon provinces (e.g.,Nigeria,Venezuela,Trinidad,Southeast Asia,the North Sea,and parts of the Middle East;Ahlbrandt et al.,2005).They have attracted considerable attention over the years with a range of publications that tackle the building of exploration-scale models (e.g.,Ainsworth et al.,2008)through to the high-resolution char-acterization of reservoirs geared toward improving field develop-ment and production management (e.g.,Chidsey et al.,2004;Hampson et al.,2008;Howell et al.,2008;Jackson et al.,2009;Ainsworth,2010).In the latter case,the detailed studies of analo-gous outcrops have helped quantify the reservoir heterogeneities,which are important in the building of subsurface models.Such studies help to highlight a hierarchy of reservoir variability imparted by shallow marine systems from field-scale architecture through to the geometry of individual sand bodies and stratal sur-faces (e.g.,Pettijohn et al.,1973;Lassiter et al.,1986;Weber,1986;Reynolds,1999;Ainsworth,2010).In recent years,workers have gone a step further and used out-crop analog data to construct detailed static models that are then dynamically simulated (Ciammetti et al.,1995;White and Barton,1999;Forster et al.,2004;Enge et al.,2007;Howell et al.,2008;Enge and Howell,2010;Jackson et al.,2009;Adams et al.,2011;Deveugle et al.,2011).Making use of the sophisticated acquisition of digital outcrop data (e.g.,LIDAR —light detection and ranging),coupled with improved reservoir modeling method-ologies and software functionality (Labourdette et al.,2008;Sech et al.,2009;Adams et al.,2011),these models provide a prediction of the impact certain reservoir heterogeneities have on connectivity and flow behavior.Many of the studies deal with modeling enhanced recovery of oil reservoirs by waterflooding andValini Gayadeen has a B.Sc.degreefrom the University of the West Indies and an M.Sc.degree from the University of Trinidad and Tobago.She has worked for BG Group since 2010as a petroleum engineer,including an assignment as an operations engineer in QGC,Australia.Helisaul J.Torrealba ∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;Helisaul.Torrealba@Helisaul Torrealba has a B.Sc.degree inpetroleum engineering from University of Oriente in Venezuela and a M.Sc.degree in petroleum engineering and project development fromIFP-School in France (Institut Francais du Petrole).He has worked for BG Group since 2011in Trinidad and Tobago.Prior to BG,he worked for Eni (Italy and Egypt),Beicip-Franlab (France,Mexico,and Brazil),and PDVSA (Venezuela).Elizabeth S.Drayton ∼BG Trinidad and Tobago Limited,BG House,5St Clair Avenue,Port of Spain,Trinidad and Tobago;Elizabeth.Drayton@Elizabeth Drayton has a B.Sc.degree from the University of Waterloo,Canada.She has worked for BG Trinidad and Tobago since 2010as a geologist where she is currently a member of the NCMA subsurface team.A C KN O WL ED G E M EN T SThe authors thank the North Coast Marine Area (NCMA)partnership:BG Group,Centrica,Eni,and Petrotrin for permission to publish this article.Although its content reflects the views of the authors,the work has benefited from technical discussion with Shyam Manrakhan,Shivanand Soloman,Arden Burrowes,Sidharth Singh,Laurent de Verteuil (now Roc Oil,Malaysia),Nigel Robertson,Leon Erriah,Marc Quesnel,Analiese Ferreira,Kat Norman,Eric Edwardson (BG Group),Roger Kimber,and Anthony Uwaga (Centrica).Considerable technical feedback and encouragement on all aspects of this work has been received from Richard Steele and Jenna Joseph (BG Group).The authors thank reviewers Anthony Reynolds,Gary Hampson,and William Hill,and AAPG Editor Michael Sweet for their constructive criticism of the original manuscript.556Dynamic Behavior of Shallow Marine Reservoirsemphasize how increased reservoir heterogeneity adversely effects sweep efficiency and ultimately reserves recovery(Kjønsvik et al.,1994;Howell et al.,2008;Enge and Howell,2010;Jackson et al., 2009;Deveugle et al.,2011).Although these models attempt to quantify the geological factors,which con-dition shallow-marine reservoir behavior,most are limited to outcrop reconstructions that do not utilize data from producing fields.Several studies,using both surface and subsurface data sets,draw attention to the importance of adopting a sequence stratigraphic approach to interwell correla-tions and the prediction of reservoir architecture (Jennette and Riley,1996;Tinker,1996;Wehr and Brasher,1996;Ainsworth et al.,1999).These studies agree with the model-based examples cited above, but also point out that sequence stratigraphic models predict more realistic reservoir architectures and there-fore potentially help define infill well targets,drive perforation strategies,and optimize reserves recovery. Although the models built in these studies make for-ward predictions on likely dynamic reservoir behavior, examples that can link historical production perfor-mance with specific aspects of reservoir heterogeneity are rare(e.g.,Cook et al.,1999).Using an integrated geological,geophysical,and petroleum engineering dataset from several Pliocene reservoirs from offshore North Trinidad,we aim to link reservoir architecture and specific hetero-geneities with historical reservoir performance and Figure1.Map of Trinidad showing the location of block NCMA-1,and the positions of Poinsettia and Chaconia(M2), Hibiscus(M4),and Bougainvillea,Heliconia,and Ixora fields (M6).Seafloor contours are also shown to indicate water depth(ft).C ROSS ET AL.557posite two-dimensional seismic line and interpretation(Jon Clemson,2009,personal communication)showing the northward-thickening Oligocene to Pleistocene stratigraphy developed along the north coast of Trinidad.The line of section(shown in Figure1)is approximately30km(∼18.6mi)northeast of the producing fields of NCMA-1.The producing Pliocene reservoirs of block NCMA-1are situated between the intra Lower Pliocene and intra Upper Miocene stratigraphic markers.Stratigraphic horizons have been dated by biostratigraphic data collected from a regional well data set that includes the well KK6-1.558Dynamic Behavior of Shallow Marine Reservoirsdemonstrate the key geological components that con-trol flow behavior in these producing gas fields. REGIONAL SETTING AND STRATIGRAPHY The North Coast Marine Area(block NCMA-1)lies 40km(25mi)off the north coast of Trinidad on the southern edge of the Caribbean tectonic plate (Escalona and Mann,2011;Garciacaro et al.,2011; Figure1).In the Trinidad region,this active plate margin exhibits dextral transform motion accommo-dated along a number of major strike-slip faults that dissect the northern(El Pilar fault zone)and central (Central Range fault zone)parts of the island (Robertson and Burke,1989;Payne,1991; Algar and Pindall,1993;Babb and Mann,1999; Garciacaro et al.,2011;Weber et al.,2011).The stratigraphy of the NCMA-1block was deposited upon a Jurassic to Cretaceous crystalline basement of igneous rocks and metasediments within the Tobago trough(Payne,1991;Escalona and Mann,2011;Garciacaro et al.,2011).To the south, basement rises and outcrops within the northern range of Trinidad as a highly deformed east–west trending zone of uplift resulting from transpressional movement along the adjacent plate margin(Ave Lallemant and Sisson,2005;Cruz et al.,2007).The Tobago trough is a wedge-shaped prism of Oligocene–Miocene to Pleistocene sediment that onlaps the basement to the south and thickens toward the Lesser Antilles high in the north(Figure2).In NCMA-1,a number of wells penetrate the basement and demonstrate a total sedimentary thickness of3 to4km(1.8to2.5mi).Basal Oligocene–Miocene sediments comprise1–1.5km(0.6–0.9mi)of mar-ginal marine to shallow marine sediments and reflect early basin deposition.Above,approximately1km ð∼0.6miÞof Pliocene–Pleistocene strata comprises shallow marine sandstones that pass offshore into progressively deeper water mudstones.The precise geometry of the Pliocene shelf is unknown due to the lack of landward wells south of NCMA-1.Regional two-dimensional seismic lines suggest it is30–40km(19–25mi)wide based on the distance from basement onlap in the south to the broad shelf edge along the northern margin of the block(Figure2).DRILLING HISTORY AND SUBSURFACE DATA IN NCMA-1The first of the Pliocene gas fields within theNCMA-1block was discovered in1975with thedrilling of the KK4-2exploration well.However,first production was not achieved until2002when a gas market had been established in Trinidadand the Hibiscus field was brought on stream.Further drilling between2006and2009resulted inthe development of six separate gas accumulations.Relatively thin,stacked reservoirs(M2,M4,and M6)have been developed by horizontalproduction wells with open-hole gravel-pack comple-tion intervals of up to2000ft(610m)in length.The M2reservoir is the producing intervalin the Chaconia and Poinsettia fields,the M4reservoir in the Hibiscus field,and the M6inthe Bougainvillea,Heliconia,and Ixora fields.Production is through the Hibiscus and Poinsettiaplatforms and consists of dry gas with an approxi-mately99%methane content.In total,blockNCMA-1,which covers an area of approximately800km2ð∼309mi2Þ,incorporates an extensive data set from9exploration and appraisal wells,16hori-zontal wells and over10yr of production data from6separate gas fields.The subsurface of NCMA-1is well imaged onthree-dimensional seismic data,originally shot in2001.The survey was shot using5000m(16,400ft)offset cables and it had an eight-streamer configura-tion with a separation of100m(328ft).The surveywas processed with bin sizes of12.5×12.5m ð41×41ftÞ,and the peak frequency through the Pliocene section is approximately60Hz.This gives a vertical resolution of approximately8mð∼26ftÞwithin the reservoirs.The gas-bearing reservoirs are particularly well imaged on seismic data due to the strong acoustic impedance contrast between the gas-bearing sands and the enveloping mudstones. In addition,the reservoirs are buried by a simple Pliocene–Pleistocene overburden that is not highly deformed.All production wells in NCMA-1are equippedwith downhole pressure gauges,and this has allowedthe dynamic performance of the reservoirs to bemonitored over time.As would be expected in aC ROSS ET AL.559producing field,wells show a decline in pressure over time as the reservoirs are depleted of gas.The rate at which this process occurs is a function of production rate,reservoir properties (e.g.,permeability),and fluid properties (in NCMA-1,gas and water).In this study,once the basic production and fluid properties were established,decline curve analysis has been used to interpret flow behavior on an interwell scale and draw conclusions relating to reservoir architec-ture as a whole.PLIOCENE RESERVOIR STRATIGRAPHY The Pliocene stratigraphy of offshore North Trinidad is more than 1km ð>0.6mi Þthick and comprises several upward-coarsening shallow marine cycles,which are interpreted as parasequences (Van Wagoner et al.,1990;Figure 3).Individual parase-quences are commonly 40to 50m (131to 164ft)thick and are characterized by a lower portion ofoffshore shelfal mudstones,siltstones,and thinly bedded sandstones,and an upper part of shelfal-to-shoreface,sand-rich strata punctuated by thin mud-stones.The M4reservoir is the only departure from this recurring motif as it has a basal interval of chan-nelized sandstones (up to 30m [98ft]thick)which overlie a significant erosion surface.On seismic sections,the Pliocene is a predomi-nantly parallel set of reflectors in which prograda-tional geometries are locally developed.Within the lower part of the Pliocene stratigraphy (below the M6reservoir),parasequences display a subtle back-stepping to aggradational stacking pattern,which is interpreted as part of a low-order transgressive systems tract (Figure 3).Above the top of the M6parasequence,sandstones become progressively higher quality,and a clear progradational stacking pattern is developed that terminates at the top of the M2sand.The exact turnaround from transgressive to regressive conditions is gradational and not defined by a single maximum flooding surface.Instead,anFigure 3.Correlation of gamma-ray log profiles across the North Coast Marine Area block NCMA-1showing the layer-cake Pliocene stratigraphy including the key reservoir horizons M2,M4,and M6in more detail.Individual reservoirs are interpreted as shelf –shoreface parasequences.The blue triangle relates to long-term transgression,the red triangle represents long-term regression.The line of corre-lation is illustrated on Figure 1.560Dynamic Behavior of Shallow Marine Reservoirsaggradational succession is suggested below the M6 horizons and defines a broad maximum flooding zone(Figure3).The scale of the individual regressive parasequences suggests that they are probably driven by fourth-order relative sea-level changes (ca.100,000yr),although biostratigraphic data are not of sufficient resolution to confirm this(Figure3). The M2,M4,and M6cycles represent the key pro-ducing intervals in NCMA-1and are the focus of this paper(Figure4).The M2Reservoir in the Chaconia and Poinsettia FieldsThe M2is an approximately60–70mð∼197–230ftÞthick heterogeneous reservoir succession in the Chaconia and Poinsettia fields.Core and well data indicate a regressive stratigraphy that is commonly built of at least two smaller scale upward-coarsening cycles across most of the NCMA-1block(Figure4). In the northern parts of the block,the M2is more sand rich and the reservoir is sharp based.Well corre-lations and seismic interpretation suggest that the M2 sand is a laterally extensive unit that was deposited across the entire NCMA-1block(Figure5,6).In the Poinsettia field sand quality diminishes on the northern flank of the field,suggesting that truncation by a number of basinward-downthrowing extensional faults occurs close to the M2shelf edge.A combina-tion of low structural relief,faulting,and the north-ward shelf-edge pinch-out of shoreface sands provides the hydrocarbon trapping mechanisms for the Chaconia and Poinsettia accumulations.Upward-coarsening cycles within the M2reser-voir grade from basal mudstone-dominated sections with occasional cm-scale,current-rippled,very fine-grained sandstones through to interbedded intervals of cm-to-dm-scale sandstones and occasional mud-stones.Sandstone beds are massive to laminated (with rare hummocky cross-stratification),and some-times upward-fining with erosional bases.Some of the thicker sandstones beds contain evidence for soft-sediment deformation.Ripple lamination is present in the finer grained sandstones,as is bioturba-tion(Planolites and small Thalassinoides).These successions grade up into thicker submeter-scale beds of fine-grained sandstone that are weakly laminated,bioclastic,and bioturbated.Sandstone porosities within this part of the reservoir average22%,with characteristic permeabilities of between20and 100mD.These upward-coarsening units are interpreted to be upward shallowing,recording the punctuated tran-sition from outer shelf mudstones-dominated deposi-tion through to progressively higher energy inner shelf and shoreface deposition.The medial,more thinly bedded sandstone intervals imply relatively rapid deposition during discrete gravity-driven events,rather than reworking by continual wave activity,and are therefore interpreted to represent inner shelf deposition below storm-weather wave base.The more thickly bedded,bioturbated units are interpreted as lower shoreface deposits.The regressive part of the reservoir is overlain by an erosionally based unit that is massively bedded with an upward-fining grain-size profile up to8m (26ft)in thickness.It comprises a glauconitic, bioclastic,and granule-rich sandstone that is locally piped into the top of the underlying regressive unit by a Glossifungites ichnofabric.The unit is interpreted as a laterally extensive blanket of sand generated by the transgressive reworking of the underlying regressive sand.In this context,the ero-sive nature of these deposits could explain the lack of more typical,cross-stratified,upper shoreface deposits within the upward-coarsening cycles below. From a reservoir quality standpoint,it is the highest quality part of the reservoir with average porosity of 28%and permeability of250mD and was the target interval for the landing of horizontal production wells during the development of the Poinsettia and Chaconia fields.Cross sections and attribute maps through the Poinsettia and Chaconia fields suggest that the architecture of the M2reservoir is complex (Figures5,6).In the Chaconia area,seismic attrib-ute extractions reveal a series of north–south-oriented channel forms several hundred meters wide,which culminate northward into lobate sand bodies.Individual lobes are up to4km(2.5mi)long and approximately3km(1.9mi)wide.Locally, the sand bodies and channels sweep round into a more northwest direction suggesting that along-shore currents actively reworked sediment acrossC ROSS ET AL.561r 562Dynamic Behavior of Shallow Marine ReservoirsFigure5.(A)Root mean square(RMS)amplitude extraction and(B)interpretation map of the M2reservoir across much of the North Coast Marine Area block NCMA-1.Gamma-ray profiles from correlative M2reservoirs are displayed on the interpretation map in red (B).The amplitude extraction is derived from a20ms window depicted on the seismic section in Figure6B.The field outlines for the Chaconia and Poinsettia fields are shown by the blue dotted outlines.The area is dissected by an east–west-oriented system of loosely linked extensional faults,which downthrow to the north and are partially responsible for closure in both fields.C ROSS ET AL.563Figure6.(A)West-to-east seismic section with(B–D)schematic interpretations and characteristic well log display through the M2res-ervoir of the Chaconia and Poinsettia fields.(E)The log profiles illustrate a composite upward-cleaning profile to the reservoir(yellow and brown fill on gamma-ray log representing sand and shale,respectively;yellow and green fill on neutron density representing pay and non-pay,respectively),with individual cycles corresponding to clinoform breaks indicated by the seismic.564Dynamic Behavior of Shallow Marine Reservoirsthe shoreface in this direction.The Chaconia field comprises at least four laterally coalesced lobes that are fed by a single channel complex(Figure5), although the reservoir is too thin for this lobe,and channel morphology is to be interpreted in detail from seismic cross section.Seismic cross sections (Figure6)suggest a lateral shingling of lobes with bounding clinoforms directed toward the west.In this area,clinoforms relate to the lateral prograda-tion of the individual lobate sand bodies.The lobate sand bodies in Chaconia are accentuated by the seis-mic amplitude response to gas-charged sandstone. Away from the Chaconia field,lower amplitudes display similar channel-fed lobes across much of the NCMA-1block.The seismic geomorphology of the M2reservoir in the Poinsettia field also exhibits lobate sand bodies.However,unlike Chaconia,individual lobes are larger,oriented in the southwesterly direction and lack the feeder channels present in Chaconia (Figure5).The lobes are characterized by abundant clinoforms that are clearly visible in both seismic maps and cross sections and again indicate a strong along-shore current direction toward the southwest. The M2clinoforms have been mapped on seismic and can be rationalized into two types:bounding clinoforms and internal clinoforms.Bounding clinoforms form the margins of the lobate sand bodies.Around the Poinsettia field,they have a strike width of approximately7kmð∼4.3miÞand a spacing of10km(6.2mi).These clinoforms are strongly arcuate in planform and represent the deposi-tional edges of large regularly spaced progradational sand bodies.In the area of the Chaconia field,they define the edges of lobate sand bodies and are on a scale and geometry similar to the bedset clinoforms described by Van Wagoner et al.,1990and inter-preted elsewhere(e.g.,O’Byrne and Flint,1995; Pattison,1995;Hampson,2000;Howell and Flint, 2004;Sømme et al.,2008).The scale and spacing of the bounding clinoforms could represent the lateral boundaries of individual parasequences.Although bounding clinoforms are clearly depicted by seismic data,they do not coincide with individual well pene-trations,and so their exact nature has to be inferred from seismic and well calibration of the internal cli-noforms subsequently described.Internal clinoforms,which are best developed through the Poinsettia area,have an along-strike (northwest to southeast)width of approximately2to 4km(∼1.2to2.5mi),and a dip(northeast to south-west)width of approximately1kmð∼0.6miÞ.Their geometry is concave-up to straight in dip profile and curvilinear to arcuate in planform.In the northern part of Poinsettia,they have a spacing of approximately 200to300m(∼656to984ft)and a depositional dip between1and2°.In Chaconia,they are not as clear in plan view as they are in cross section,in which they have a similar preferential orientation to the southwest.Figure6displays the calibration between the log and seismic character of internal cli-noforms.The seismic character is driven by the acoustic impedance contrast between interbedded sand and mudstone,and this is accentuated within the toesets of upward-cleaning cycles.Stacked upward-cleaning cycles(with a scale of40m [131ft]),which build the larger scale regressive motif of the M2reservoir correspond to the well-defined clinoforms exhibited by seismic cross sections (Figure6).Although a shoreface-to-shelf interpretation can be derived by core sedimentology,seismic geomor-phology and attribute mapping provide more detail on sediment transport and larger scale processes that controlled the deposition of the M2reservoir.The partially channelized flows and associated sediment lobes that traversed the block in a northward direction (e.g.,toward the Chaconia field)suggests a relatively defuse line source of channels with a spacing of4to 5km(2.5to3.1mi)rather than a single major deltaic point of deposition.The predominance of channels with terminal lobes implies a significant fluvial con-tribution of sand to the shoreface.Seismic cross sec-tions also show that,although sediment transport was dominantly from the south,westerly directed alongshore currents resulted in the compensational stacking of lobes in this direction(Chaconia field). As a result,seismic cross sections through the Chaconia area show a similar east-to-west orientation of clinoforms as those present in Poinsettia (Figures5,6).In the Poinsettia area,east-to-west migrating lobes with abundant internal clinoforms testify to long-shore-current-driven bars that migrated throughC ROSS ET AL.565。