# Introduction to Monte Carlo Analysis Consulting in Oil and Gas Reserve Estimation

Various Types of Models

Numerical Models

All numerical models have input values.

Examples are discrete numbers such as 1, 2, 350 or continuous such as 2.2354532 or 1 and 7/8ths.

The value may be absolutely certain or stochastic (random pattern).

Stochastic input values are usually used in models, rarely ever have absolute certainty.

Stochastic Models

The input values will follow any one of the numerous statistical distributions, for example the Normal or Gaussian distribution or the Uniform distribution.

Example; human population height follows a normal distribution.

Selection of which distribution depends on scientific observation (geologists) of historical data and professional judgment.

Limitation of Stochastic Models

Stochastic models by their very nature cannot be calculated definitively, unlike say the square footage of your house (length x width = total area).

Stochastic models do provide the average answer (assuming that all input values represent the average input value) but tell you NOTHING of the range or probability of all possible answers.

This range of possibility is critical when determining the profitability of a venture, safety of a drug or integrity of a building.

Monte Carlo Analysis Consulting Method

A method utilizing random sampling of the Stochastic input values to provide a picture of the output distribution values and their probabilities.

The quality of the random number generator http://www.greckoglobal.com/oil-gas - Monte Carlo Simulation in Oil and Gas - is critical, Crystal Ball used the Multiplicative Congruential Generator algorithm which is multitudes better than the excel random number generator.

One iteration randomly samples each stochastic input, utilizing a new set of random input values from each possible distribution.

The model then takes these inputs and utilizing the algorithm calculates the respective outputs

These singular outputs are recorded and finally summated.

The process is repeated x times (usually 10,000) until sufficient repeat samples are collected to provide a probability breakdown for a range of output values.

Monte Carlo Simulation in Oil and Gas Example Reserve Estimation

Calculation of potential oil reserves.

Limited information available of extent of reserve, rock type, pressure, gas content, water content, and percentage hydrocarbons.

Utilize the Monte Carlo Method to bracket uncertainty.

Hydrocarbon in Place = GRV x N/G x Porosity x Hydrocarbon in Place = GRV x N/G x Porosity x

Sh / FVF

Where:

Gross Rock Volume - amount of rock in the trap above the hydrocarbon water contact.

N/G - net/gross ratio net/gross ratio - percentage of the GRV http://www.greckoglobal.com/oil-gas - Oil and Gas Consulting - formed by the reservoir rock (range is 0 to 1).

Porosity - percentage of the net reservoir rock occupied by pores (typically 5 pores (typically 5-35%).

Sh - hydrocarbon saturation hydrocarbon saturation - some of the pore space is filled with water - this must be discounted.

FVF - formation volume factor formation volume factor - oil shrinks and gas expands when brought to the surface. The FVF converts volumes at reservoir conditions (high pressure and high temperature) to storage and sale conditions.

Recoverable Hydrocarbons

Recoverable Hydrocarbons = Hydrocarbons in Place x Percentage Recoverable Hydrocarbons.

Recoverable hydrocarbons - amount of hydrocarbon likely to be recovered during production. This is typically 10-50% in an oil field and 50-80% in a gas field.

How can the results be used?

Use Oil and Gas Consulting to determine whether rewards of extraction outweigh the risks.

This can be done economically if one can add cost of extraction & transportation with expected price of oil to the model then calculate the range of revenues and profits together with probabilities.

Comparisons can be made with other oil extraction options the company may have to determine most likely http://www.greckoglobal.com/oil-gas - Oil and Gas Consulting - ROI (productive field).

Numerical Models

All numerical models have input values.

Examples are discrete numbers such as 1, 2, 350 or continuous such as 2.2354532 or 1 and 7/8ths.

The value may be absolutely certain or stochastic (random pattern).

Stochastic input values are usually used in models, rarely ever have absolute certainty.

Stochastic Models

The input values will follow any one of the numerous statistical distributions, for example the Normal or Gaussian distribution or the Uniform distribution.

Example; human population height follows a normal distribution.

Selection of which distribution depends on scientific observation (geologists) of historical data and professional judgment.

Limitation of Stochastic Models

Stochastic models by their very nature cannot be calculated definitively, unlike say the square footage of your house (length x width = total area).

Stochastic models do provide the average answer (assuming that all input values represent the average input value) but tell you NOTHING of the range or probability of all possible answers.

This range of possibility is critical when determining the profitability of a venture, safety of a drug or integrity of a building.

Monte Carlo Analysis Consulting Method

A method utilizing random sampling of the Stochastic input values to provide a picture of the output distribution values and their probabilities.

The quality of the random number generator http://www.greckoglobal.com/oil-gas - Monte Carlo Simulation in Oil and Gas - is critical, Crystal Ball used the Multiplicative Congruential Generator algorithm which is multitudes better than the excel random number generator.

One iteration randomly samples each stochastic input, utilizing a new set of random input values from each possible distribution.

The model then takes these inputs and utilizing the algorithm calculates the respective outputs

These singular outputs are recorded and finally summated.

The process is repeated x times (usually 10,000) until sufficient repeat samples are collected to provide a probability breakdown for a range of output values.

Monte Carlo Simulation in Oil and Gas Example Reserve Estimation

Calculation of potential oil reserves.

Limited information available of extent of reserve, rock type, pressure, gas content, water content, and percentage hydrocarbons.

Utilize the Monte Carlo Method to bracket uncertainty.

Hydrocarbon in Place = GRV x N/G x Porosity x Hydrocarbon in Place = GRV x N/G x Porosity x

Sh / FVF

Where:

Gross Rock Volume - amount of rock in the trap above the hydrocarbon water contact.

N/G - net/gross ratio net/gross ratio - percentage of the GRV http://www.greckoglobal.com/oil-gas - Oil and Gas Consulting - formed by the reservoir rock (range is 0 to 1).

Porosity - percentage of the net reservoir rock occupied by pores (typically 5 pores (typically 5-35%).

Sh - hydrocarbon saturation hydrocarbon saturation - some of the pore space is filled with water - this must be discounted.

FVF - formation volume factor formation volume factor - oil shrinks and gas expands when brought to the surface. The FVF converts volumes at reservoir conditions (high pressure and high temperature) to storage and sale conditions.

Recoverable Hydrocarbons

Recoverable Hydrocarbons = Hydrocarbons in Place x Percentage Recoverable Hydrocarbons.

Recoverable hydrocarbons - amount of hydrocarbon likely to be recovered during production. This is typically 10-50% in an oil field and 50-80% in a gas field.

How can the results be used?

Use Oil and Gas Consulting to determine whether rewards of extraction outweigh the risks.

This can be done economically if one can add cost of extraction & transportation with expected price of oil to the model then calculate the range of revenues and profits together with probabilities.

Comparisons can be made with other oil extraction options the company may have to determine most likely http://www.greckoglobal.com/oil-gas - Oil and Gas Consulting - ROI (productive field).

## Replies