## Engineering variables The following input data are required, but some are optional, in which case the program makes assumptions on the basis of formula that often depend on the depth of the reservoir:

For these data the input form is the UDD on each line of the table.

The temperature gradient and the surface temperature are required to calculate subsurface temperatures. In turn, these are involved in the calculation of a number of reservoir engineering parameters:

• Gas/Oil ratio (GOR)
• Formation Volume Factor
• Z-factor (Gas compressibility)
• Expansion factor
• In situ oil density
• In situ gas density

### Surface Temperature

Note that in the administration part of the input the waterdepth is required. On land this would be taken to be zero, but in the offshore, or in a lake, it is important to take the surface temperature to be the sea- or lake-bottom temperature. The reason is that at the sea bottom considerably lower temperatures can prevail than on land. The gradient is applied only in the subsurface below sea bottom, hence a more accurate estimate of the subsurface temperature can be obtained. The surface temperature is the annual mean temperature, so depends on the latitude. For lake/sea bottom temperature, the bottom temperature can be estimated with the latitude and the waterdepth as shown in the following graph, given by Beardsmore & Cull, 2001. The gradient is expressed in degrees C per 100m (or degrees Fahrenheit/1000 ft in OFUs) and the surface/bottom temperature in degrees C (or degrees Fahrenheit).

Worldwide the gradient is on the order of 3 �C/100m (16.455 �F/1000ft). This is also the mean default value used by the program if no input is given. The default for the surface/bottom temperature is 10 �C. The reservoir temperature estimation uses the T-gradient and applies this to the depth difference between either surface or waterbottom and the reference depth of the reservoir. So the increase of temperature with depth takes place in the rock-overburden only. The latter is estimated, not at the culmination of a trap, but at a point representative of the center of gravity of the volume of trapped HC, if any.

### Reservoir Pressure

The formation pressure of the target reservoir is a function of the water density in the overburden. Normally we can assume a standard gradient of the usually salty water in the overburden. However there is a water density input or a water salinity input possible. If these are used the pressure gradient is calculated from these values, otherwise a default gradient based on 1020 kg/M� water density is used.

Pressure is an optional input. If not entered, it is assumed that the section is hydrostatic and the subsurface pressure is calculated accordingly. Deviations from the hydrostatic can be due to presence of a HC column in a reservoir. Then gas columns usually will create a differential pressure at the top of the column that is representing the buoyancy of gas on water. An oil column, being heavier, usually has less effect, but also creates differential pressure. Exception is very heavy oil. Pressure input in Gaeapas is the pressure estimate in the absence of HC!
The graph of pressure versus depth shows various types of deviation of formation pressure from the hydrostatic. Note that the curve does not start at pressure=0 at the surface, but at 1 bar (= the pressure of the earth atmosphere). In other words, calculations are on the basis of "absolute pressures". The input becomes important if the target formation is suspected to be overpressured. In the extreme case the pressure is some 2.5 times the hydrostatic: the whole overburden weight is on the formation water. This is usually referred to as "lithostatic".

On form 11.1 the pressure input can be made directly in the UDD inpit line. You may also have some data available from other reservoirs in the area, which are overpressured. The button "Analyse analogon pressure data" opens a window where data pairs of pressure and depth can be entered. When done, an estimate of the pressure in the target reservoir is calulated from such data. The algorithm assumes (somewhat arbitrarily) that the data are all lying on a single overpressure trend versus depth.

The input is in terms of bar (one bar is a "hectopascal" or 105 Pa) or PSI ("Pounds per Square Inch"). In a hydrostatic section the pressure increases at a rate of about one bar per ten meter. A bar is slightly different from the older atmosphere unit.

### Gas/Oil Ratio

The gas oil ratio (GOR) is expressed in cubic meter per cubic meter, or in OFU units as cubic feet per barrel. It is the ratio of the dissolved gas over liquid oil. In condensate reservoirs the GOR is sometimes measured as the ratio of the gas produced and the volume of condensate associated with it. Then very high GOR values can result. For normal conditions with liquid oil at reservoir conditions, the GOR will largely be limited to about 5000 cuft/b, or about 900 M3/M3 as a maximum. The input refers to the subsurface reservoir conditions, but both the oil and gas volumes are taken at standard conditions. GOR input should be the "Solution GOR" (sometimes denoted as Rs). In most cases it is misleading to take the ratio of produced gas over produced oil from a well ("Production GOR") as a guide! Better data are obtained from the "PVT" analysis ("Pressure/Volume/Tempareture").

In the entrapment phase the PVT calculations are made to assess the volume of oil and gas in a trap at subsurface conditions.

Under the Contacts option the user can also make a GOR input (Gas/Oil Ratio). The effect of this is that only when the Gas/Oil contact (GOC) is equal to the culmination the user GOR input is used. In all other cases the Vasquez-Beggs equations are used to estimate GOR.

When looking at the expectation curves of the GOR, the mean of the individual GOR values in the MC cycles is not normally corresponding to the ratio of standard condition Solution gas over standard conditions Oil. This is particularly the case in the FMB option. The reason is the the mean of ratios is not the same as the ratio of the means. Which value to use in further analysis depends on the purpose.

### Oil gravity or density

This is an optional input. If not used then a worldwide calibration of oil density on depth is used by the program. The prior empirical distribution is based on a sample of 370 densities from 47 basins.

Density is in kg/m� or in API; the probable range for inputs is between 700 and 1000 (Note that in SI terms the density of pure water is 1000 kg/m� and not expressed as 1 g/cm�).

To the right of the oil density/gravity input boxes is a yellow button "Bayesian update". When clicked, this will show a form where up to 20 oil density/gravity data can be entered. Other input requirements are the depth at which the density was observed and the waterdepth at that location.

1. The prior world-wide data are almost exclusively from onshore oil fields.
2. Oil density is more related to subsurface temperature than depth.
3. The temperature gradient is relevant only to the "rock overburden".
4. The analog sample data have to be made equivalent to the situation of the prospect to be evaluated. A shift is made parallel to the regression curve to the prospect depth. This means calculating the residual for the sample data, as shown in the figure below, but also using the waterdepth of the prospect to be evaluated.
5. The sample data are assumed to be normally distributed. Mean and satndard deviation are calculated.
6. Having also the data from the world-wide regression, the conditional density distribution for the prospect can be calculated, using its reference depth. This is the prior normal distribution.
7. Bayes theorem for the conjugate normal distribution gives the posterior normal distribution required. This is a weighted combination of the sample information and the prior world-wide information. The procedure is the Bayes normal update.
8. A normal posterior distribution might generate physically impossible values for the density (e.g. negative density). Therefore the normal distribution is trimmed to the best fitting triangular distribution. This is the suggested input. The "Accept and insert" button puts the three numbers in the three input boxes for oil density/gravity.  The effect of a few "equalized" API sample data (blue) on the worldwide prior (white) distribution. The red posterior probability density function is the required basis for suggesting an input.

### Formation Volume Factor

This is the ratio ("FVF") of the subsurface volume of oil to that at surface standard conditions. It is usually a number between 1 and 6. The factor is expressed as volume/volume. So it would be the same whether Metric or OFU. Input can be made in three cells, allowing a constant, rectangular or triangular distribution.

If left blank, the calibrated subroutine calculates the FVF value as the default. This calculation uses the Gas/Oil ratio, apart from other variables (Vasquez-Beggs correlation). The GOR, in turn, may be user supplied or default (also Vasquez-Beggs). If both GOR and FVF are user input, their relationship should be realistic for the reservoir PVT. If the GOR is default ( no user input) then a user-defined FVF might well be in conflict with the GOR. In principle, leaving both factors to Vasquez-Beggs is therefore preferable and assures a consistent result. But Vaqsquez-Beggs may not always give acceptable results in case the GOR and FVF are already known, when evaluating a discovery already made.

The FVF is also often refrerred to as Boi where the B stands for "barrel", the subsript "o" for oil and the "i" for initial conditions in the reservoir when discovered. So, oil "shrinks" when it is brought to the surface. The "shrinkage factor" is 1/FVF. Similarly there is the term Bg that is referring to gas, but then it is the reciprocal of the expansion factor.

It is useful to know that oil "swells" in the subsurface by dissolving gas, thus increasing the FVF (or Boi) while free gas is compressed at depth. Gas expands if brought to the surface.

### Expansion factor

The Expansion Factor is the ratio of the volume of gas at surface standard conditions to that in the reservoir. A rough guide to its value is to take the depth in meter of the reservoir and divide by 10 to get the expansion factor.

The expansion factor can be calculated using the Boyle-GayLussac formula for the volume of gas at standard conditions and for the sub-surface conditions and dividing the first by the second: Where:
V is the volume
R is the universal gasconstant (8314.51 Joule/KMol)
n is the number of moles of gas
P is the absolute pressure (Bar)
T is the absolute temperature (Degrees Kelvin)
Z is the gas deviation factor.Assumed here Zsc= 1.0 at standard conditions.
E is the expansion factor
sc subscript for "standard conditions" of Tsc=290,
and Psc=1.

Another term used in this context is Bg that is referring to gas, but then it is the reciprocal of the expansion factor.

### Condensate/gas ratio, or richness

The condensate richness is a parameter that is automatically calculated on a calibrated basis. The correlations of this parameter with the various geological generation parameters is not very strong. Therefore the user may prefer his own estimates, that may be based on cases in the same play.

An important question to answer is "What is condensate". Here we consider condensate as HC that are in condenste phase form in the free gas in the subsurface. Only PVT analyis may indicate this. However, at the well-head the characteristics of the HCs produced do not give a clear answer about the subsurface phase behaviour. Cova, et al, 1992 has established some rules for areas in Venezuela, which may be indicative in general as well. A GOR of above 445 M�/M� (2500 CuFt/barrel) and liquid densities of less than 825 Kg/M� (API 40) would be a fairly certain indicator for a condensate phase in the reservoir. However at less GOR and higher densities also a condensate phase may occur.

Because it is difficult to predict whether liquids forecasted by Gaeapas are black oil or condensate, it may be that at greater reservoir depth, it is likely that what is shown as oil, should actually be regarded as condensate. The "Total Liquids" in the tables and curves may then be more appropriate than the separate oil and condensate numbers.

The input is in terms of m� liquid per million m� of free gas. If there is input, it will override the calibrated estimation.

The final amount of recoverable condensate is calculated using this richness factor on the free gas in place and then multiplying with the condensate recovery fraction.

### Gas composition and non-hydrocarbon gas

An optional input. If not given, the program assumes 100 percent HC gas. The user is asked to estimate the percentage of non-HC gas, which usually is the non-combustible part of the gas. This affects the usefulness ("calorific value") of the gas as a fuel. For instance, the gas from the Groning field (Netherlands onshore) contains 15.2% non-combustible gas, mainly Nitrogen and a bit of CO2.

The usual admixtures in natural gas are N2, CO2, H2S, and He. If the non-HC gas content is high, it may be worth removing these components to obtain combustible gas with a calorific value. This comes at a considerable investment and operating cost.
The non-combustible gas admixtures would influence the Z-factor estimates, but this program does not take that into account.

The geochemical processes that generate the non-combustible components are not fully understood, or not to the level of giving a guide to calibration. H2S can be produced from the sulphur in kerogen by bacteria when reservoir temperature is less than 80 �C (176 �F). Geothermal processes play an important role at higher temperatures, say 150 to 250 �C ( ~300-500 �F). See Nagihara,2005. N2 and CO2 can be the result of high temperatures and volcanism.
Helium makes up from a fraction of a percent to 10% in extreme cases. It comes in two forms: He-4 and He-3. He-4 is the result of radioactive decay, while the rarer He-3 is mainly an original constituant of the earth's mantle.

A good survey of non-HC gas in the NW Europe and the possible geochemical processes involved is given by Gras & Clayton (1998).

### World-wide a priory distributions

A calibration of recovery efficiency was made by GAEA on the basis of published data on existing fields. The prime purpose of establishing a database containing recovery efficiency and reservoir characteristics was to find relationships. However, no significant correlations were observed in this data set (85 reservoirs in 60 basins). Hence, only world-wide prior distributions were obtained. These form WW-default inputs that can be invoked by the user, in case he has no better information available to him. These are for oil, free gas and condensate, both primary and primary+secondary recovery. In this exercise no data were available to subdivide the fields according to drive mechanisms. Studies in the US have shown that knowledge of the drive mechanism, such as waterdrive, depletion drive, etc, would narrow the predictive distribution.

In-place volumes of HC

The output summary table with Mean Success Volumes and expectations for the various HC types is in terms of recoverable volumes only. If the user requires in place volumes (e.g. "STOIIP") the recovery efficiency must be set at 100% and the program re-run. But in place volumes are available as condensed expectation curves in the View/Summary/Condensed ECs, but not as a tabulation.

Primary and Secondary recovery

The first oil recovery input refers to the primary recovery efficiency for oil, i.e. the percentage of the oil in place that can be recovered by primary methods. This may involve various natural drive mechanisms, such as waterdrive or depletion drive. The menu item Input/WW defaults are worldwide triangular distribution parameters for primary recovery or primary+secondary recovery. These are based on a calibration of recovery efficiency of more than 100 fields worldwide, notably those described in the AAPG treatise of petroleum geology.

The user is free to change these and when using Primary+secondary, or enhanced recovery methods, he should make a note that the economic evaluation should include the additional costs and investments that such methods would require. If not, the evaluation would be inbalanced.

Automatic default recovery efficiency

The World-wide default distributions are automatically used if the recovery inputs are left empty!

### Gas recovery efficiency

This first UDD for recovery input refers again to primary recovery efficiency of free gas. The second is for Primary+secondary recovery.The discussion under oil recovery efficiency holds equally for this parameter. This program assumes that the recovery efficiency of solution gas is the same as that of the oil. The user should be aware of this simplification if solution gas is a significant contributor to value of the prospect.

The world-wide default distributions are automatically used if the recovery inputs are left empty!

### Condensate recovery

Two UDD input controls specify primary condensate recovery and Primary+secondary. Condensate recovery is a complicated subject. Normally the condensate comes with the free gas that is produced. Here the user will have to consider the technical possibilities carefully. The tendency is that the higher the condensate richness, the lower the efficiency may become. This can be caused by retrograde condensation of the condensate in the gascap, thereby blocking the pores for gas production.

Recycling of gas is a common method to produce the condensate at high efficiencies. So, here again, expensive installations are associated with the higher recoveries and the economic evaluation should take this into account.

### User Economic Cutoff Volumes

On the right side of the input grid, three input boxes allow the user to specify an economic minimum volume of the resource for economic development, the cutoff. There is a cutoff for oil, total gas and condensate. If these boxes are left blank the cutoffs are assumed to be zero.