You must log in to edit PetroWiki. Help with editing

Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information


Production forecasting glossary: Difference between revisions

PetroWiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
Line 1: Line 1:
This glossary was created through discussions among the steering committee for the SPE Global Integrated Workshop Series (GIWS) on Production Forecasting. Some definitions were not contested at all, others generated fierce discussions.
This glossary was created through discussions among the steering committee for the SPE Global Integrated Workshop Series (GIWS) on Production Forecasting. Some definitions were not contested at all, others generated fierce discussions.


== A ==


==A==
'''ACQ'''
'''ACQ'''


Line 9: Line 9:
'''Availability'''
'''Availability'''


A factor applied to forecasts to take into account the fact that a Production System will not always operate at 100% of its capacity.  
A factor applied to forecasts to take into account the fact that a Production System will not always operate at 100% of its capacity.


See also: IPSC and ABNR
See also: IPSC and ABNR
Line 17: Line 17:
Available But Not Required, that part of the IPSC that is available for production but not produced because of low off-take demand. ABNR typically occurs in Demand Driven Production Systems.
Available But Not Required, that part of the IPSC that is available for production but not produced because of low off-take demand. ABNR typically occurs in Demand Driven Production Systems.


==C==
== C ==


'''Choke Model'''
'''Choke Model'''


A model visualising and quantifying a series of restrictions or chokes that represent the areas of production loss, i.e. Reservoir potential (as measured by well tests), that is not, or cannot be, delivered at custody transfer point. These are simplified for management reporting as Reservoir, Wells, Plant, and Export chokes.  
A model visualising and quantifying a series of restrictions or chokes that represent the areas of production loss, i.e. Reservoir potential (as measured by well tests), that is not, or cannot be, delivered at custody transfer point. These are simplified for management reporting as Reservoir, Wells, Plant, and Export chokes.


'''Constraint'''
'''Constraint'''


A production or injection constraint is a factor limiting production or injection capacity. This can be a physical constraint induced by existing or planned equipment capacity (e.g. compressor capacity) or an operational constraint induced by commercial, political, social, economic, regulatory, permitting, government, or environmental guidelines (e.g. flare limits, max allowable reservoir voidage, contractual production limits, etc.).  
A production or injection constraint is a factor limiting production or injection capacity. This can be a physical constraint induced by existing or planned equipment capacity (e.g. compressor capacity) or an operational constraint induced by commercial, political, social, economic, regulatory, permitting, government, or environmental guidelines (e.g. flare limits, max allowable reservoir voidage, contractual production limits, etc.).


'''Controllable variable'''
'''Controllable variable'''
Line 33: Line 33:
'''CTP'''
'''CTP'''


Custody Transfer Point, the point at which the forecast volumes are quantified.  
Custody Transfer Point, the point at which the forecast volumes are quantified. Note: this may be different from the volumes produced from the reservoir and wells, as it takes into account “own use”, flare/vent volumes and potential losses in the production system.
Note: this may be different from the volumes produced from the reservoir and wells, as it takes into account “own use”, flare/vent volumes and potential losses in the production system.


[[#Top|Back to top]]  
[[#Top|Back to top]]


==D==
== D ==


'''DCQ'''
'''DCQ'''
Line 50: Line 49:
'''Deferment, or Deferred Production/Injection'''
'''Deferment, or Deferred Production/Injection'''


Deferment is the reduction in Production or Injection
Deferment is the reduction in Production or Injection Availability caused by an activity, breakdown, trip, poor equipment performance, or sub-optimum operations, that results in a reduction in the volume sold or injected, delaying the production or injection until some later time.
Availability caused by an activity, breakdown, trip, poor equipment performance, or sub-optimum operations, that results in a reduction in the volume sold or injected, delaying the production or injection until some later time.


See also: Scheduled/Unscheduled Deferments.
See also: Scheduled/Unscheduled Deferments.
Line 57: Line 55:
'''Demand Driven Production System'''
'''Demand Driven Production System'''


A Production System with Load Factor potentially < 1.  
A Production System with Load Factor potentially < 1. Production from a Demand Driven System is determined by external product demand. This applies mainly to gas production systems where gas demand may fluctuate.
Production from a Demand Driven System is determined by external product demand. This applies mainly to gas production systems where gas demand may fluctuate.  


'''Demand Profile'''
'''Demand Profile'''
Line 66: Line 63:
'''Deterministic Forecast'''
'''Deterministic Forecast'''


A deterministic forecast is combining a single set of discrete parameter estimates (gross rock volume, average porosity, development plan, equipment availability, etc.) that represent a physically realizable and realistic combination in order to derive a single, specific estimate of production per time step and/or recoverable quantities.  
A deterministic forecast is combining a single set of discrete parameter estimates (gross rock volume, average porosity, development plan, equipment availability, etc.) that represent a physically realizable and realistic combination in order to derive a single, specific estimate of production per time step and/or recoverable quantities.


A deterministic forecast is based on a specific combination of parameters representing a specific scenario, but does not give quantitative information on the probability of a certain outcome, even if the probability of the individual input parameters is known.
A deterministic forecast is based on a specific combination of parameters representing a specific scenario, but does not give quantitative information on the probability of a certain outcome, even if the probability of the individual input parameters is known.
Line 72: Line 69:
'''Dynamic Reservoir Simulation Model'''
'''Dynamic Reservoir Simulation Model'''


A dynamical system is a concept in mathematics where a fixed rule describes the time dependence of a point in a geometrical space.
A dynamical system is a concept in mathematics where a fixed rule describes the time dependence of a point in a geometrical space. A dynamic reservoir simulation model is a computer model used to predict the flow of fluids in a reservoir.
A dynamic reservoir simulation model is a computer model used to predict the flow of fluids in a reservoir.


==E==
== E ==


'''EoFL'''
'''EoFL'''


End of Field Life, a point in time when production from a field ended, for instance when abandonment conditions have been reached.  
End of Field Life, a point in time when production from a field ended, for instance when abandonment conditions have been reached.


'''Expectation Forecast'''
'''Expectation Forecast'''
Line 85: Line 81:
See Mean Forecast
See Mean Forecast


[[#Top|Back to top]]  
[[#Top|Back to top]]


==F==
== F ==


'''FDP'''
'''FDP'''
Line 95: Line 91:
'''Forecast'''
'''Forecast'''


In general: a forecast is a prediction based on study and/or observation.
In general: a forecast is a prediction based on study and/or observation. In the Oil & Gas Industry, a forecast is the projection in time of a quantity (production, injection, emissions, energy consumption, etc.); the projection is computed using a quantitative model that processes input data.
In the Oil & Gas Industry, a forecast is the projection in time of a quantity (production, injection, emissions, energy consumption, etc.); the projection is computed using a quantitative model that processes input data.  


==H==
== H ==


'''High Estimate (resources)'''
'''High Estimate (resources)'''
Line 108: Line 103:
Modifying reservoir model to match the historical production and pressure data with simulated results as closely as possible. The history matching quality depends on the quality of the reservoir model and the quality and quantity of historical data. The history matched reservoir models are subsequently used to help predict future reservoir behaviour.
Modifying reservoir model to match the historical production and pressure data with simulated results as closely as possible. The history matching quality depends on the quality of the reservoir model and the quality and quantity of historical data. The history matched reservoir models are subsequently used to help predict future reservoir behaviour.


==I==
== I ==


'''IAM'''
'''IAM'''
Line 118: Line 113:
'''IPSC'''
'''IPSC'''


Integrated Production System Capacity, the capacity the integrated production system is, or would be able to deliver at the Custody Transfer Point for the integrated production system at 100% availability.
Integrated Production System Capacity, the capacity the integrated production system is, or would be able to deliver at the Custody Transfer Point for the integrated production system at 100% availability. Note: This is not necessarily equal to the capacity that could be produced from wells at 100% availability as it is defined at delivery point and takes into account own use and vent/flare volumes.
Note: This is not necessarily equal to the capacity that could be produced from wells at 100% availability as it is defined at delivery point and takes into account own use and vent/flare volumes.


IPSC = Production + Deferment (+ Losses) + ABNR
IPSC = Production + Deferment (+ Losses) + ABNR
Line 129: Line 123:
Integrated Production System Model, modelling the flow and changes in composition of hydrocarbons and other fluids through wells and facilities from the reservoir to delivery point, taking into account the effects of (re)injection, flare, vent, “own use”, surface treatments.
Integrated Production System Model, modelling the flow and changes in composition of hydrocarbons and other fluids through wells and facilities from the reservoir to delivery point, taking into account the effects of (re)injection, flare, vent, “own use”, surface treatments.


[[#Top|Back to top]]


[[#Top|Back to top]]
== L ==
 
==L==


'''L - M - H'''
'''L - M - H'''


Low - Medium –High, three discrete values to describe the uncertainty range of forecast input parameters, whereby the medium value is the "best estimate".  
Low - Medium –High, three discrete values to describe the uncertainty range of forecast input parameters, whereby the medium value is the "best estimate".


L - M -H can also be used to describe the uncertainty range for deterministic forecasts, in which case no probability may be attached to the individual scenarios.
L - M -H can also be used to describe the uncertainty range for deterministic forecasts, in which case no probability may be attached to the individual scenarios.
Line 146: Line 139:
'''Load Factor'''
'''Load Factor'''


The ratio of the average load to the peak load during any particular time period. The Load Factor is a measure of the continuity of the demand. A Load Factor less than 1 means that one will not be able to sell the total volume that the system could physically deliver at the custody transfer point.
The ratio of the average load to the peak load during any particular time period. The Load Factor is a measure of the continuity of the demand. A Load Factor less than 1 means that one will not be able to sell the total volume that the system could physically deliver at the custody transfer point.


'''Long Term'''
'''Long Term'''


The period from now until end-of-field-life (EOFL). LT forecasts are generally prepared for resource volume estimates and expressed in annual volumes.  
The period from now until end-of-field-life (EOFL). LT forecasts are generally prepared for resource volume estimates and expressed in annual volumes.


'''Losses'''
'''Losses'''
Line 160: Line 153:
With respect to resource categorization, this is considered to be a conservative estimate of the quantity that will actually be recovered from the accumulation by a project. If probabilistic methods are used, there should be at least a 90% probability (P90) that the quantities actually recovered will equal or exceed the low estimate.
With respect to resource categorization, this is considered to be a conservative estimate of the quantity that will actually be recovered from the accumulation by a project. If probabilistic methods are used, there should be at least a 90% probability (P90) that the quantities actually recovered will equal or exceed the low estimate.


==M==
== M ==


'''Mean Forecast'''
'''Mean Forecast'''


In probabilistic production forecasting: the concatenation in time of mean values of the forecast range as calculated per time-step.  
In probabilistic production forecasting: the concatenation in time of mean values of the forecast range as calculated per time-step. Not necessarily the forecast that results from using the mean values per uncertain input parameter!
Not necessarily the forecast that results from using the mean values per uncertain input parameter!


Note: Expectation forecasts may be aggregated (Sum of the Mean is the Mean of the Sum)
Note: Expectation forecasts may be aggregated (Sum of the Mean is the Mean of the Sum)
Line 171: Line 163:
'''Medium Term'''
'''Medium Term'''


The period starting now and ending one or two years beyond the start of the next budget year. This period is defined separately from ST (in-year) and LT (EoFL) because the quality of this forecast is much dependent on the contributions and ownership from various Disciplines (e.g. Development, Operations). The MT forecast is affected by Controllable Variables such as Opex/Capex investments and equipment performance/availability.
The period starting now and ending one or two years beyond the start of the next budget year. This period is defined separately from ST (in-year) and LT (EoFL) because the quality of this forecast is much dependent on the contributions and ownership from various Disciplines (e.g. Development, Operations). The MT forecast is affected by Controllable Variables such as Opex/Capex investments and equipment performance/availability. The MT forecast may be expressed in terms of annual volumes or (average) daily rates.
The MT forecast may be expressed in terms of annual volumes or (average) daily rates.


'''Model'''
'''Model'''
Line 178: Line 169:
A (set of) mathematical equation(s) that converts (model) input parameters to (model) output parameters, given certain inner and outer boundary conditions.
A (set of) mathematical equation(s) that converts (model) input parameters to (model) output parameters, given certain inner and outer boundary conditions.


[[#Top|Back to top]]  
[[#Top|Back to top]]


==N==
== N ==


'''NFA'''
'''NFA'''
Line 196: Line 187:
'''NTR'''
'''NTR'''


Non Technical Risk.  
Non Technical Risk. These are risks directly affecting a specific project, caused by (non-contractor) external Stakeholders that trigger a deviation from the locally established and/or expected behaviours / practices / regulations. What sets NTR apart from technical risks within a project context is that there is a clear linkage to external stakeholders, i.e. parties in commercial, political, social, economic, regulatory, permitting, government, and environmental areas that have a real or perceived interest in the project.
These are risks directly affecting a specific project, caused by (non-contractor) external Stakeholders that trigger a deviation from the locally established and/or expected behaviours / practices / regulations.
What sets NTR apart from technical risks within a project context is that there is a clear linkage to external stakeholders, i.e. parties in commercial, political, social, economic, regulatory, permitting, government, and environmental areas that have a real or perceived interest in the project.


==O==
== O ==


'''Own Use'''
'''Own Use'''


Produced hydrocarbons used to operate the system, such as: Continuous Flare pilot, (Continuous) Vent purge, Fuel (e.g. for power generation and compressors), Blanket Gas.  
Produced hydrocarbons used to operate the system, such as: Continuous Flare pilot, (Continuous) Vent purge, Fuel (e.g. for power generation and compressors), Blanket Gas. Own Use gas is produced gas not available for sales.
Own Use gas is produced gas not available for sales.


[[#Top|Back to top]]  
[[#Top|Back to top]]




==P==
 
== P ==


'''Planned Deferments'''
'''Planned Deferments'''


The allowance made in the forecast for Scheduled and Unscheduled Deferments, usually expressed as a % of IPSC.
The allowance made in the forecast for Scheduled and Unscheduled Deferments, usually expressed as a&nbsp;% of IPSC. Note: If a forecast contains an allowance for Unscheduled Deferments, this is called Planned Unscheduled Deferments.
Note: If a forecast contains an allowance for Unscheduled Deferments, this is called Planned Unscheduled Deferments.


'''Probabilistic'''
'''Probabilistic'''


Quantifying a range of values per uncertain input parameter. The range of values per uncertain input parameter is described by a probability distribution.
Quantifying a range of values per uncertain input parameter. The range of values per uncertain input parameter is described by a probability distribution. Quantifying the uncertainty of a prediction can only be done probabilistic ally.
Quantifying the uncertainty of a prediction can only be done probabilistic ally.  


'''Probability'''
'''Probability'''


The extent to which an event is likely to occur, measured by the ratio of the favourable cases to the whole number of cases possible. SPE convention is to
The extent to which an event is likely to occur, measured by the ratio of the favourable cases to the whole number of cases possible. SPE convention is to quote cumulative probability of exceeding or equalling a quantity where P90 is the small estimate and P10 is the large estimate. (See also Uncertainty.)
quote cumulative probability of exceeding or equalling a quantity where P90 is the small estimate and P10 is the large estimate. (See also Uncertainty.)


'''Probabilistic Forecast'''
'''Probabilistic Forecast'''


A probabilistic forecast is the result from sampling from ranges of values per uncertain input parameter (gross rock volume, average porosity, development plan, equipment availability, etc.) that each represent a physically realizable and realistic combination in order to derive a range of estimates and associated probabilities of production per time step and/or recoverable quantities.  
A probabilistic forecast is the result from sampling from ranges of values per uncertain input parameter (gross rock volume, average porosity, development plan, equipment availability, etc.) that each represent a physically realizable and realistic combination in order to derive a range of estimates and associated probabilities of production per time step and/or recoverable quantities.


See also: Deterministic Forecast.
See also: Deterministic Forecast.
Line 235: Line 221:
'''Production Forecast'''
'''Production Forecast'''


Forward looking estimate of future saleable production of hydrocarbon products per time step, with the objective to quantify the rate or volume per time step. Time steps can be yearly, monthly or daily, depending on the length of the forecast: LT, MT, ST.  
Forward looking estimate of future saleable production of hydrocarbon products per time step, with the objective to quantify the rate or volume per time step. Time steps can be yearly, monthly or daily, depending on the length of the forecast: LT, MT, ST.


See also: Recovery Forecast
See also: Recovery Forecast
Line 241: Line 227:
'''P50 Production Forecast'''
'''P50 Production Forecast'''


A (probabilistic) Pxx Production Forecast refers to the probability of achieving the production rates or volumes per time step, NOT the probability of achieving a specified cumulative recovery.
A (probabilistic) Pxx Production Forecast refers to the probability of achieving the production rates or volumes per time step, NOT the probability of achieving a specified cumulative recovery. The P50 Production Forecast can only be prepared probabilistically. For a series of time steps, the P50 Production Forecast is the concatenation of P50 values per time step. This does NOT represent a specific realisation or scenario! See also: P50 Recovery Forecast Notes&nbsp;: 1) Because of the skewed distribution of these volumes per time step, the aggregation of P50 values per time step over the lifetime of a field does NOT lead to the P50 UR value. 2) There are many different forecasts that lead to the P50 UR value. 3) A full probabilistic forecast would provide both the ranges of uncertainty per time step as well as the range of uncertainty in UR.
The P50 Production Forecast can only be prepared probabilistically. For a series of time steps, the P50 Production Forecast is the concatenation of P50 values per time step. This does NOT represent a specific realisation or scenario!
See also: P50 Recovery Forecast
Notes :
1) Because of the skewed distribution of these volumes per time step, the aggregation of P50 values per time step over the lifetime of a field does NOT lead to the P50 UR value.
2) There are many different forecasts that lead to the P50 UR value.
3) A full probabilistic forecast would provide both the ranges of uncertainty per time step as well as the range of uncertainty in UR.


'''Production (volume)'''
'''Production (volume)'''
Line 255: Line 235:
'''Production Forecast Quality'''
'''Production Forecast Quality'''


The quality of a single production forecast cannot be established due to the fundamentally probabilistic nature of the forecasting problem. The quality can only be established for a production forecasting method, i.e. following repeated experiments that have been generated according to this method.  
The quality of a single production forecast cannot be established due to the fundamentally probabilistic nature of the forecasting problem. The quality can only be established for a production forecasting method, i.e. following repeated experiments that have been generated according to this method. The quality of the method is then revealed by the question whether learning is enabled, i.e. whether an objective function, comparing actual vs. forecast, is consistently being improved over the life-cycle of the asset. The function should enable systematic learning, i.e. (1) increased accuracy (i.e., the mean value of the ensemble of stochastic realizations of forecasts is increasingly close to actual); (2) increased precision (i.e., the range of the ensemble of stochastic realizations is increasingly narrow); (3) increased time horizon (i.e., longer periods for which criteria 1 and 2 are valid).
The quality of the method is then revealed by the question whether learning is enabled, i.e. whether an objective function, comparing actual vs. forecast, is consistently being improved over the life-cycle of the asset.
The function should enable systematic learning, i.e.  
(1) increased accuracy (i.e., the mean value of the ensemble of stochastic realizations of forecasts is increasingly close to actual);  
(2) increased precision (i.e., the range of the ensemble of stochastic realizations is increasingly narrow);  
(3) increased time horizon (i.e., longer periods for which criteria 1 and 2 are valid).


See also: Objective Function
See also: Objective Function


[[#Top|Back to top]]


[[#Top|Back to top]]
== R ==
 
==R==


'''Recovery Forecast'''
'''Recovery Forecast'''


Forward looking estimate of future saleable production of hydrocarbon products per time step, covering the period until EoFL with the objective to quantify the total recoverable volume from a resource.  
Forward looking estimate of future saleable production of hydrocarbon products per time step, covering the period until EoFL with the objective to quantify the total recoverable volume from a resource.


See also: Production Forecast and P50 Recovery Forecast
See also: Production Forecast and P50 Recovery Forecast
Line 279: Line 253:
A Pxx Recovery Forecast is a (deterministic) forecast scenario leading to the quantified Pxx UR volume. It quantifies the probability of achieving the cumulative value of forecast “ultimate recovery”, NOT the probability per time step. Therefore, the time when the UR may be achieved (EoFL) remains uncertain, only the volume that may be recovered at EoFL is quantified.
A Pxx Recovery Forecast is a (deterministic) forecast scenario leading to the quantified Pxx UR volume. It quantifies the probability of achieving the cumulative value of forecast “ultimate recovery”, NOT the probability per time step. Therefore, the time when the UR may be achieved (EoFL) remains uncertain, only the volume that may be recovered at EoFL is quantified.


A P50 Recovery Forecast is a forecast for which the cumulative production at EoFL equals the P50 UR. The probability of achieving the P50 UR should not be confused with the P50 probability of achieving a volume per time step. The latter is depicted by the P50 Production Forecast.
A P50 Recovery Forecast is a forecast for which the cumulative production at EoFL equals the P50 UR. The probability of achieving the P50 UR should not be confused with the P50 probability of achieving a volume per time step. The latter is depicted by the P50 Production Forecast. In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.
In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.


In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.
In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.


[[#Top|Back to top]]


[[#Top|Back to top]]
== S ==
 
==S==


'''Scenario'''
'''Scenario'''


A deterministic definition of one or more uncertain forecast input parameters. A scenario features as a discrete branch of a chance node in a decision tree, and has a probability of occurrence associated with it.  
A deterministic definition of one or more uncertain forecast input parameters. A scenario features as a discrete branch of a chance node in a decision tree, and has a probability of occurrence associated with it. A scenario can also include a selection of controllable variables (decisions made at t=0).
A scenario can also include a selection of controllable variables (decisions made at t=0).


A scenario is not a stochastic realisation of a model output.
A scenario is not a stochastic realisation of a model output.
Line 306: Line 277:
'''Supply Driven Production System'''
'''Supply Driven Production System'''


A Production System with Load Factor = 1.  
A Production System with Load Factor = 1. Production from a Supply Driven System is determined by the available IPSC. This applies mainly to oil production systems where production is maximised.
Production from a Supply Driven System is determined by the available IPSC. This applies mainly to oil production systems where production is maximised.


==T==
== T ==


'''Technical Risks'''
'''Technical Risks'''
Line 321: Line 291:
An event with a probability of occurrence in the time period considered of less than 10% and a magnitude sufficient to move the P50 of the forecast distribution outside of the P10/90 range. Events with a probability lower than 1% are usually ignored.
An event with a probability of occurrence in the time period considered of less than 10% and a magnitude sufficient to move the P50 of the forecast distribution outside of the P10/90 range. Events with a probability lower than 1% are usually ignored.


[[#Top|Back to top]]


[[#Top|Back to top]]
== U ==
 
==U==


'''(Contractual) Under-deliveries'''
'''(Contractual) Under-deliveries'''


The difference between the demand and the sales if demand is higher than sales. The calculation of the contractual under delivery depends on the sales/delivery contract.
The difference between the demand and the sales if demand is higher than sales. The calculation of the contractual under delivery depends on the sales/delivery contract.


'''Unplanned Deferments'''
'''Unplanned Deferments'''


The actual Deferments (Scheduled and Unscheduled) for which no allowance was made in the forecast, usually expressed in volume terms or as % of actual IPSC.
The actual Deferments (Scheduled and Unscheduled) for which no allowance was made in the forecast, usually expressed in volume terms or as&nbsp;% of actual IPSC.


'''Unscheduled Deferments'''
'''Unscheduled Deferments'''
Line 338: Line 307:
Deferment is referred to as Unscheduled when it is caused by events that cannot be scheduled (e.g. trips, breakdowns and interruptions by external parties), poor performance of equipment or sub-optimum settings.
Deferment is referred to as Unscheduled when it is caused by events that cannot be scheduled (e.g. trips, breakdowns and interruptions by external parties), poor performance of equipment or sub-optimum settings.


[[#Top|Back to top]]  
[[#Top|Back to top]]
 








==References==
== References ==
 
Use this section for citation of items referenced in the text to show your sources. [The sources should be available to the reader, i.e., not an internal company document.]
Use this section for citation of items referenced in the text to show your sources. [The sources should be available to the reader, i.e., not an internal company document.]


==Noteworthy papers in OnePetro==
== Noteworthy papers in OnePetro ==
 
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read


==External links==
== External links ==
 
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro


Line 363: Line 336:


[http://petrowiki.org/Sandbox:Production_forecasting_purpose Sandbox:Production forecasting purpose]
[http://petrowiki.org/Sandbox:Production_forecasting_purpose Sandbox:Production forecasting purpose]
== Category ==
[[Category:5.6 Formation evaluation and management]] [[Category:5.6.9 Production forecasting]] [[Category:GIWS-PF]]

Latest revision as of 16:29, 31 March 2016

This glossary was created through discussions among the steering committee for the SPE Global Integrated Workshop Series (GIWS) on Production Forecasting. Some definitions were not contested at all, others generated fierce discussions.

A

ACQ

Annual Contractual Quantity. The contract quantity is the contractually agreed volumes and limits: predefined (annual) volume of natural gas on contract level.

Availability

A factor applied to forecasts to take into account the fact that a Production System will not always operate at 100% of its capacity.

See also: IPSC and ABNR

ABNR

Available But Not Required, that part of the IPSC that is available for production but not produced because of low off-take demand. ABNR typically occurs in Demand Driven Production Systems.

C

Choke Model

A model visualising and quantifying a series of restrictions or chokes that represent the areas of production loss, i.e. Reservoir potential (as measured by well tests), that is not, or cannot be, delivered at custody transfer point. These are simplified for management reporting as Reservoir, Wells, Plant, and Export chokes.

Constraint

A production or injection constraint is a factor limiting production or injection capacity. This can be a physical constraint induced by existing or planned equipment capacity (e.g. compressor capacity) or an operational constraint induced by commercial, political, social, economic, regulatory, permitting, government, or environmental guidelines (e.g. flare limits, max allowable reservoir voidage, contractual production limits, etc.).

Controllable variable

A variable that can be influenced by staff through improved working procedures, for example down-time. See also: non-controllable variable.

CTP

Custody Transfer Point, the point at which the forecast volumes are quantified. Note: this may be different from the volumes produced from the reservoir and wells, as it takes into account “own use”, flare/vent volumes and potential losses in the production system.

Back to top

D

DCQ

Daily Contracted Quantity. Can be absolute or derived from the ACQ.

Decline Curve Analysis

A method of forecasting well or reservoir production by extrapolating past trends in well or reservoir behaviour.

Deferment, or Deferred Production/Injection

Deferment is the reduction in Production or Injection Availability caused by an activity, breakdown, trip, poor equipment performance, or sub-optimum operations, that results in a reduction in the volume sold or injected, delaying the production or injection until some later time.

See also: Scheduled/Unscheduled Deferments.

Demand Driven Production System

A Production System with Load Factor potentially < 1. Production from a Demand Driven System is determined by external product demand. This applies mainly to gas production systems where gas demand may fluctuate.

Demand Profile

A demand profile is a time series that describes the expected future demand over time for a given counterparty.

Deterministic Forecast

A deterministic forecast is combining a single set of discrete parameter estimates (gross rock volume, average porosity, development plan, equipment availability, etc.) that represent a physically realizable and realistic combination in order to derive a single, specific estimate of production per time step and/or recoverable quantities.

A deterministic forecast is based on a specific combination of parameters representing a specific scenario, but does not give quantitative information on the probability of a certain outcome, even if the probability of the individual input parameters is known.

Dynamic Reservoir Simulation Model

A dynamical system is a concept in mathematics where a fixed rule describes the time dependence of a point in a geometrical space. A dynamic reservoir simulation model is a computer model used to predict the flow of fluids in a reservoir.

E

EoFL

End of Field Life, a point in time when production from a field ended, for instance when abandonment conditions have been reached.

Expectation Forecast

See Mean Forecast

Back to top

F

FDP

Field Development Plan

Forecast

In general: a forecast is a prediction based on study and/or observation. In the Oil & Gas Industry, a forecast is the projection in time of a quantity (production, injection, emissions, energy consumption, etc.); the projection is computed using a quantitative model that processes input data.

H

High Estimate (resources)

With respect to resource categorization, this is considered to be an optimistic estimate of the quantity that will actually be recovered from an accumulation by a project. If probabilistic methods are used, there should be at least a 10% probability (P10) that the quantities actually recovered will equal or exceed the high estimate.

History Matching

Modifying reservoir model to match the historical production and pressure data with simulated results as closely as possible. The history matching quality depends on the quality of the reservoir model and the quality and quantity of historical data. The history matched reservoir models are subsequently used to help predict future reservoir behaviour.

I

IAM

Integrated Asset Model

See also Glossary:IAM

IPSC

Integrated Production System Capacity, the capacity the integrated production system is, or would be able to deliver at the Custody Transfer Point for the integrated production system at 100% availability. Note: This is not necessarily equal to the capacity that could be produced from wells at 100% availability as it is defined at delivery point and takes into account own use and vent/flare volumes.

IPSC = Production + Deferment (+ Losses) + ABNR

See also: Availability

IPSM

Integrated Production System Model, modelling the flow and changes in composition of hydrocarbons and other fluids through wells and facilities from the reservoir to delivery point, taking into account the effects of (re)injection, flare, vent, “own use”, surface treatments.

Back to top

L

L - M - H

Low - Medium –High, three discrete values to describe the uncertainty range of forecast input parameters, whereby the medium value is the "best estimate".

L - M -H can also be used to describe the uncertainty range for deterministic forecasts, in which case no probability may be attached to the individual scenarios.

LNG

Liquified Natural Gas

Load Factor

The ratio of the average load to the peak load during any particular time period. The Load Factor is a measure of the continuity of the demand. A Load Factor less than 1 means that one will not be able to sell the total volume that the system could physically deliver at the custody transfer point.

Long Term

The period from now until end-of-field-life (EOFL). LT forecasts are generally prepared for resource volume estimates and expressed in annual volumes.

Losses

That part of the IPSC that could have been delivered to customers but is not, due to leaks, theft, etc.

Low Estimate (resources)

With respect to resource categorization, this is considered to be a conservative estimate of the quantity that will actually be recovered from the accumulation by a project. If probabilistic methods are used, there should be at least a 90% probability (P90) that the quantities actually recovered will equal or exceed the low estimate.

M

Mean Forecast

In probabilistic production forecasting: the concatenation in time of mean values of the forecast range as calculated per time-step. Not necessarily the forecast that results from using the mean values per uncertain input parameter!

Note: Expectation forecasts may be aggregated (Sum of the Mean is the Mean of the Sum)

Medium Term

The period starting now and ending one or two years beyond the start of the next budget year. This period is defined separately from ST (in-year) and LT (EoFL) because the quality of this forecast is much dependent on the contributions and ownership from various Disciplines (e.g. Development, Operations). The MT forecast is affected by Controllable Variables such as Opex/Capex investments and equipment performance/availability. The MT forecast may be expressed in terms of annual volumes or (average) daily rates.

Model

A (set of) mathematical equation(s) that converts (model) input parameters to (model) output parameters, given certain inner and outer boundary conditions.

Back to top

N

NFA

No Further Activity: the forecast without new Capex invested.

Nomination

Nomination is a binding commitment or order to deliver, buy or transport gas.

Non-controllable Variable

A variable that cannot be influenced by staff through improved working procedures, for example permeability.

NTR

Non Technical Risk. These are risks directly affecting a specific project, caused by (non-contractor) external Stakeholders that trigger a deviation from the locally established and/or expected behaviours / practices / regulations. What sets NTR apart from technical risks within a project context is that there is a clear linkage to external stakeholders, i.e. parties in commercial, political, social, economic, regulatory, permitting, government, and environmental areas that have a real or perceived interest in the project.

O

Own Use

Produced hydrocarbons used to operate the system, such as: Continuous Flare pilot, (Continuous) Vent purge, Fuel (e.g. for power generation and compressors), Blanket Gas. Own Use gas is produced gas not available for sales.

Back to top


P

Planned Deferments

The allowance made in the forecast for Scheduled and Unscheduled Deferments, usually expressed as a % of IPSC. Note: If a forecast contains an allowance for Unscheduled Deferments, this is called Planned Unscheduled Deferments.

Probabilistic

Quantifying a range of values per uncertain input parameter. The range of values per uncertain input parameter is described by a probability distribution. Quantifying the uncertainty of a prediction can only be done probabilistic ally.

Probability

The extent to which an event is likely to occur, measured by the ratio of the favourable cases to the whole number of cases possible. SPE convention is to quote cumulative probability of exceeding or equalling a quantity where P90 is the small estimate and P10 is the large estimate. (See also Uncertainty.)

Probabilistic Forecast

A probabilistic forecast is the result from sampling from ranges of values per uncertain input parameter (gross rock volume, average porosity, development plan, equipment availability, etc.) that each represent a physically realizable and realistic combination in order to derive a range of estimates and associated probabilities of production per time step and/or recoverable quantities.

See also: Deterministic Forecast.

Production Forecast

Forward looking estimate of future saleable production of hydrocarbon products per time step, with the objective to quantify the rate or volume per time step. Time steps can be yearly, monthly or daily, depending on the length of the forecast: LT, MT, ST.

See also: Recovery Forecast

P50 Production Forecast

A (probabilistic) Pxx Production Forecast refers to the probability of achieving the production rates or volumes per time step, NOT the probability of achieving a specified cumulative recovery. The P50 Production Forecast can only be prepared probabilistically. For a series of time steps, the P50 Production Forecast is the concatenation of P50 values per time step. This does NOT represent a specific realisation or scenario! See also: P50 Recovery Forecast Notes : 1) Because of the skewed distribution of these volumes per time step, the aggregation of P50 values per time step over the lifetime of a field does NOT lead to the P50 UR value. 2) There are many different forecasts that lead to the P50 UR value. 3) A full probabilistic forecast would provide both the ranges of uncertainty per time step as well as the range of uncertainty in UR.

Production (volume)

Production is the cumulative quantity of petroleum that has been actually recovered over a defined time period. While all recoverable resource estimates and production are reported in terms of the sales product specifications, raw production quantities (sales and non-sales, including non-hydrocarbons) are also measured to support engineering analyses requiring reservoir voidage calculations.

Production Forecast Quality

The quality of a single production forecast cannot be established due to the fundamentally probabilistic nature of the forecasting problem. The quality can only be established for a production forecasting method, i.e. following repeated experiments that have been generated according to this method. The quality of the method is then revealed by the question whether learning is enabled, i.e. whether an objective function, comparing actual vs. forecast, is consistently being improved over the life-cycle of the asset. The function should enable systematic learning, i.e. (1) increased accuracy (i.e., the mean value of the ensemble of stochastic realizations of forecasts is increasingly close to actual); (2) increased precision (i.e., the range of the ensemble of stochastic realizations is increasingly narrow); (3) increased time horizon (i.e., longer periods for which criteria 1 and 2 are valid).

See also: Objective Function

Back to top

R

Recovery Forecast

Forward looking estimate of future saleable production of hydrocarbon products per time step, covering the period until EoFL with the objective to quantify the total recoverable volume from a resource.

See also: Production Forecast and P50 Recovery Forecast

P50 Recovery Forecast

A Pxx Recovery Forecast is a (deterministic) forecast scenario leading to the quantified Pxx UR volume. It quantifies the probability of achieving the cumulative value of forecast “ultimate recovery”, NOT the probability per time step. Therefore, the time when the UR may be achieved (EoFL) remains uncertain, only the volume that may be recovered at EoFL is quantified.

A P50 Recovery Forecast is a forecast for which the cumulative production at EoFL equals the P50 UR. The probability of achieving the P50 UR should not be confused with the P50 probability of achieving a volume per time step. The latter is depicted by the P50 Production Forecast. In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.

In other words: There is a difference between the P50 Production Forecast (depicting probability of achieving a certain rate or volume per time step) and the P50 Recovery Forecast (depicting a particular and realistic scenario leading to P50 UR, but not quantifying the probability of achieving a certain rate or volume per time step). The P50 value of reserves is not equal to the cumulative production resulting from the P50 Production Forecast. However, a full probabilistic forecast could provide both the range of uncertainty in recovery volume as well as the ranges of uncertainty of produced volumes per time step.

Back to top

S

Scenario

A deterministic definition of one or more uncertain forecast input parameters. A scenario features as a discrete branch of a chance node in a decision tree, and has a probability of occurrence associated with it. A scenario can also include a selection of controllable variables (decisions made at t=0).

A scenario is not a stochastic realisation of a model output.

Scheduled Deferment

Deferment is referred to as Scheduled when it is caused by activities that have or should have been planned and scheduled. A scheduled activity or event has a defined start and end point.

Short Term

Any period less than one year or within the current year. ST forecasts are generally prepared to minimise deferments and/or contractual under deliveries, optimise activity scheduling and to provide an update of expected production for the running year (Latest Estimate).

Supply Driven Production System

A Production System with Load Factor = 1. Production from a Supply Driven System is determined by the available IPSC. This applies mainly to oil production systems where production is maximised.

T

Technical Risks

Technical Risks are those issues which are closer to the project and generally within the project teams direct ability to influence (eg. contracting or permitting / approvals within the project’s sphere of influence), or are common to the delivery environment of the project (ie. the country approvals mechanism, or weather and other “repeat” environmental issues).

See also: Non Technical Risks

Train Wreck

An event with a probability of occurrence in the time period considered of less than 10% and a magnitude sufficient to move the P50 of the forecast distribution outside of the P10/90 range. Events with a probability lower than 1% are usually ignored.

Back to top

U

(Contractual) Under-deliveries

The difference between the demand and the sales if demand is higher than sales. The calculation of the contractual under delivery depends on the sales/delivery contract.

Unplanned Deferments

The actual Deferments (Scheduled and Unscheduled) for which no allowance was made in the forecast, usually expressed in volume terms or as % of actual IPSC.

Unscheduled Deferments

Deferment is referred to as Unscheduled when it is caused by events that cannot be scheduled (e.g. trips, breakdowns and interruptions by external parties), poor performance of equipment or sub-optimum settings.

Back to top



References

Use this section for citation of items referenced in the text to show your sources. [The sources should be available to the reader, i.e., not an internal company document.]

Noteworthy papers in OnePetro

Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read

External links

Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro

See also

Sandbox:Production forecasting building blocks

Sandbox:Production forecasting expectations

Sandbox:Production forecasting flowchart

Sandbox:Production forecasting in the financial markets

Sandbox:Production forecasting purpose

Category