Data Analysis

Slide 11

• Here is an example of a structural trap – a simple anticline
• On the left is a map view; right is cross-section A-A'
• If enough HC has migrated into the trap, it will be filled to a spill or leak point For this example, the spill point is on the ENE
• Any more HC added will spill to the ENE, possibly filling another trap further along the migration pathway

Slide 12

• If the amount of HC reaching the trap is small (less than the trap could hold), then the trap is said to be under-filled
• In this example the trap is ‘HC charge-limited’ and is not filled to the spill point
• The way the cross-section is drawn, both ‘bumps’ have been filled to about the same level
• This would be true if HC migration came from the NW or SE
• If HC migration came from the west, it would first fill the western ‘bump’ to the saddle point and then HC would start to fill the eastern ‘bump’
• If HC migration came from the east, how would the two ‘bumps’ fill? Fill east first; then spill to west

Slide 13

• Here is another type of structural trap with two degrees of fill
• On the left, cross-section on top and map view below
  • where the reservoir on the down side touches the fault is the leak point
  • more HC fill would leak across and up the fault plane and move through the high-side sands
• On the right, cross-section on top and map view below
  • here the fault 'seals' – does not allow HC to penetrate it
  • the reservoir fill level is controlled by a synclinal leak point on the west
• This illustrates the importance of predicting whether faults leak or seal HCs
• It could be that the left case does not hold enough HC to be an economic success, but the right case does hold enough HC to be an economic success
• There are methods to predict the sealing potential of faults, but that topic is beyond the scope of this course

Slide 14

• Here are two stratigraphic traps
• On the left, cross-section on top and map view below
  • This represents a sub-unconformity style of trap
  • sands below the unconformity have a component of dip and form the reservoir
  • above the unconformity is a marine shale that provides the seal
  • here the big unknown often is how far downdip the sands contain HCs
• On the right, cross-section on top and map view below
  • This represents an isolated reef
    • porous and permeable carbonate facies form the reservoir – often in the edges of the reef rather than the central lagoonal facies
    • Again there has to be an adequate seal facies capping the reef

Slide 15

• This slide illustrates one combination trap
• Top left = structure map with a high in the center
• Top right = depositional facies – a deep water channel system where the axial facies has the best reservoir properties
• Bottom right = cross-section showing structure, depositional facies, and green is the oil fill
• Bottom left = map view with structure contours, facies belts and oil within reservoir quality rocks

• The anticinal structure causes oil to migrate towards the high point (crest)
• The facies belts limit the producible reservoir to the channel axis facies
• The trap has both a stuctural & a stratigraphic component

Slide 16

Next we will highlight some geophysical evidence for the presence of HCs

There are two types of evidence:

• DHIs – Direct HC Indicators
• AVO – Amplitude Versus Offset

Slide 17

A DHI is a direct hydrocarbon indicator

• This means that there is something anomalous in the seismic response caused by the presence of HCs
• When the porous rock has HC in the pore spaces of the rock, the combination of velocity and density results in a diagnostic seismic anomaly (a DHI)

There are a number of DHI signatures, such as:

• an Amplitude anomaly
• a Fluid contact reflection
• the anomaly exhibits a good Fit to structure

Slide 18

The graph is based in data from the Gulf of Mexico – impedance versus depth for shales and for sands with 3 types of fluids in the pores: water, oil, gas

Look around 4,000 ft

• shale and water sand have about the same impedance so reflection amplitudes would be weak
• oil sand has an impedance that is about 15% lower
  • if a thick oil sand is encased above and below by shales and water sands, it would be associated
    • with moderate to high reflection amplitudes (top and base, opposite polarities)
• gas sand has an impedance that is about 40% lower
  • if a thick gas sand is encased above and below by shales and water sands, it would be associated
    • with extremely high reflection amplitudes (top and base, opposite polarities)
• at 4,000 ft we should be able to differentiate oil sands from gas sands by the amplitude response

• At 7,500 feet, an oil sand would have about 8% lower impedance; gas about 25% lower
  • there would still be significant impedance contrasts, but the amplitudes would not be as strong
• The deeper the potential reservoir, the less diagnostic the amplitude response
• Gas sand would be diagnostic; oil sands might be harder to detect

• Using reflection amplitudes to detect the presence of HCs is commonly referred to as "Bight Spot" technology
• It works in many basins, not just the Gulf of Mexico

Slide 19

• These images are from some modeled seismic data with noise included
• The left image shows two regions with anomalous amplitudes – dark blacks and big excursions in the troughs (whites)
• In the upper region, the basal strong reflection also appears rather flat – another clue

• In the right image, the modeled reservoir is thicker
• There are several clues to note:
  • the reflection at the top (yellow line) changes from low to high amplitude as it goes from a water sand to a HC sand
  • the strong basal reflection on the right indicates the bottom of the HC-filled sand
  • towards the center, the strong black becomes nearly flat
  • this flat to dipping geometry at the base of the HC-bearing zone is called a "hockey stick"
  • the central (flat) portion is a reflection off the fluid contact – HC-filled sand above with water-filled sand below
  • although the sand properties are the same, having different fluids results in different velocities and densities for the formation
• So on the right we have two lines of evidence
  • an amplitude anomaly at the top and base of the HC-bearing rock, and
  • a fluid contact reflection

Slide 20

• Here are two more seismic models (with added noise) to illustrate a fluid contact reflection
• Note the "hockey stick" in the upper image
• For the lower image, the reservoir is thinner we only see the fluid contact, not the handle portion of the "hockey stick"

• IMPORTANT – a fluid contact will be flat (horizontal) in depth, unless there is a hydrodynamic gradient causing it to tilt
• Since we are looking at data in the time domain, there can be some distortions if the velocities above the reservoir
  • vary laterally and cause a flat event in time to be tilted on a time section

Slide 21

• If we have a prospect with a fluid reflection, that fluid reflection should be flat in depth
• We can plot where the fluid reflection is and post it on a map
• We can then measure how well it conforms to a constant depth around a structure (anticline)
• If we are working in units of two-way time, it usually show a good to fair fit to structure
• If the fit to structure is not good, we would convert the interpreted top of reservoir horizon from two-way time to depth (ft or meters) and check again
• It is rare that the fluid contact actually dips in depth due to a hydrodynamic gradient

Slide 22

• AVO is an analysis of how reflection amplitude varies as a function of incident angle
• I lied before when I said the reflection coefficient was a function of velocities and densities – please forgive me
• There is another factor that can be important, the angle of incident of the seismic energy hitting the interface
• If a shot and receiver are close, then the incident angle is close to 90°
• However, if the shot-to-receiver distance is several kilometers, the incident angle could be greater than 30°
• The impact of the incident angle is different for shales, water-sands, oil-sands and gas-sands
• AVO analysis capitalizes on these differences

• We can ask our data processors to stack the data (combine shot-receiver pairs) in different ways
  • A full stack uses all shot-receiver pairs at all incident angles
  • A near stack might use only the receivers in the first (nearest to the boat) half of the streamers
  • A far stack might use only the receivers in the second (farthest from the boat) half of the streamers
• We can then compare the reflection amplitude from a reflection off the top of a potential reservoir from the near stack relative to that on the far stack

Slide 23

• This shows one source-receiver pair and illustrates the incidence angle Θ

• Note: the physics dictates that we really want Θ, the angle of incidence
• It is easier for us to stack the seismic data by distance along the streamer/cable
• AVO is amplitude versus offset – or distance along the streamer
• AVA is amplitude versus angle – a better way to do things but a bit more expensive
• Many companies do AVA in the data processing, but get sloppy in their terminology and call it AVO analysis

Slide 24

• This slides explains (again) why we care about AVA or AVO
• Note in the diagram, the top of the reservoir is in a trough
  • The excursion to the left (amplitude) for the near offset stack is noticeably less than the excursion to the left (amplitude) for the far offset stack
• The same can be said for the base of the reservoir and its associated peak (black)

• Why the polarity change (top = trough, base = peak)?
  • The yellow HC sand probably has a lower impedance than the other layers
  • So there is a decrease in impedance at the top of the reservoir (higher to lower impedance) and an increase in impedance at the base of the reservoir (lower to higher impedance)

Slide 25

• We need a way to easily analyze amplitude changes with angle or offset
• Geophysicists defined two parameters for us: A and B
• The diagram on the left is a flattened CDP gather – the common traces for a location with the 'hyperbolas' removed by correcting the gather with an appropriate velocity
  • the trace on the left is close to 90° incidence; the one on the right is close to 35°
• The red line is the position of the top of a reservoir

• The second diagram shows the amplitude of the trough with angle (or offset)
  • it is in a trough so the amplitudes are negative
  • the values become more negative with angle (or offset)
• We can approximate the amplitude variation with a straight line fit
  • this straight line can be characterized by a slope (gradient) and an intercept (value when x = 0; 90° incidence)
• The parameter A is the intercept value; B is the slope (or gradient)
• We can express the amplitude vs offset as a point on the right chart, which plots intercept (x axis) against slope (y axis)

• Typically we build seismic models and fill a sand layer with the three types of fluid – water, oil, gas
• Each fluid type will lie in a different portion of the A B crossplot (right chart)
• The more separate the three points, the easier it is to differentiate water-sands from oil-sands from gas-sands

Slide 26

• Here is a seismic line with good DHI evidence for HCs
• The top of reservoir reflection changes amplitude with depth indicating a possible change in fluid type
• There is a good fluid reflection at the oil-water contact; a more subtle one at the gas-oil contact
• When we perform an AVO analysis, on an A versus B crossplot we get a three clouds of amplitude points that indicate we have gas high on the structure, oil at intermediate depths, and water-filled sands below the lower fluid event

Slide 27

Our interpretation shows us the present-day structure and stratigraphy From this we can:

• Predict where Sand Fairways & Source Intervals occur by predicting EODs and inferring lithologies
• Evaluate Trap Configurations by identifying and sizing potential traps and considering spill / leak points
• Consider if sealing units exist by determining if shales or other low permeability lithologies provide top & lateral seal
• Identify where a distinct HC response occurs through DHI and/or AVO analysis
• Model a simple HC migration case by using present-day dips on stratal units and assuming buoyancy-driven migration

Slide 28

• We would like to know more by including the element of geologic time
• When did the traps form?
• When did the source rocks generate HCs?
• What was the attitude (dip) of the strata when the HCs were migrating?
• What is the quality of the reservoir (Φ , k) ?
• How adequate is the seal?
• How have temperature and pressure conditions changed through time?

To answer these questions, we have to model the basin's history from the time of deposition to the present

Slide 29

That brings us to topic #5 Basin modeling

There are two approaches:

• to go back in time from the present-day configurations – geohistory
• to model the basin's development forward through geology time - simulation

Slide 30

• Here is a simple cross-section with 4 major depositional units
• brown, orange, yellow and green
• We can back-strip each unit and estimate the basin's shape – unit by unit
• We are limited to the ages of the interpreted horizons – here 18 Ma, 29 Ma, 36 Ma and before deposition began at 42 Ma

• After we do the back-stripping, we can estimate the parameters that controlled the basin's development,
  • such as subsidence through time, sediment supply through time, paleobathymetry, and some estimate of sea level change through time
• We can then use basin simulation software to model the basin forward through time
• We can use a much smaller time step, e.g., ½ million year
• We can also model temperatures, pressures, HC generation, porosity changes, etc

Slide 31

Recap of how we might model a basin's history

Slide 32

• This animation shows output from a HC migration software package
• Buoyancy is the driving mechanism and rock properties have to be defined
• In this example, Trap A has a very good seal – it has early oil fill which is displaced by late gas
  • Trap B has a good seal for oil but allows gas to leak once the gas thickness exceeds a critical value
  • Trap B maintains an oil leg – the gas can not displace all the oil since it leaks through the top seal
  • Excess oil in Trap B (with some dissolved gas) spills into Trap C

• This is a 2-D model to help show conceptually the types of HC migration that could occur
• Rock properties can be varied (e.g., stronger to weaker seals, different source rock properties)
• We can build 3-D migration models as well, but often we do not have adequate data to constrain the results
• We model different scenarios, but can not say which is correct pre-drill

Slide 33

• This introduces the exercise
• You will be mapping part of the A1 sand – the part where gas is in the pore space
• Note the strong (high negative amplitude) trough
• Also note there are faults
  • On this line a fault forms the western boundary of the gas field

Slide 34

• You will be looking at the magnitude of the trough – its largest negative amplitude
• The blow-up on the left shows extremely negative numbers
  • where the trough is vertical, it has been clipped for display purposes
• Compare this response to the blow-up on the right
• On the right, where water is the pore fluid, the response is weaker (less negative numbers)

Slide 35

You can using the trough amplitudes to predict the fluid within the sand, as shown for inline 840