Resources
The oil and gas industry needs to make forecasts of production, revenues, and expenses for a multi-year period in the face of much uncertainty. Find below three examples to deal with this type of problem using @RISK:
1. A deterministic model to get started.
2. A basic @RISK model with a number of uncertainties.
3. A version using the RiskCollect function to help with a sensitivity analysis of the basic @RISK model.
A model that illustrates how the Time Series Batch Fit feature can be used to fit several potentially correlated series all at once.
A model that illustrates the Time Series Fit feature can be used to generate future forecasts.
A set of examples to illustrate a volumetric reserves calculation in the oil industry:
1. A deterministic model to get started.
2. A basic @RISK model where the output, a product of three uncertain quantities, is shown to be lognormally distributed.
3. A version that uses several potential distributions of uncertain inputs to compare their effects on the output distribution.
4. A version where the amount of methane is the product of five uncertain quantities.
5. This model is similar to other volumetric analyses in this series, but the context now is the Austin Chalk horizontal well, as reported in a Journal of Petroleum Technology article. There are now five uncertain inputs that are multiplied or divided to obtain the Reserves output.
In many circumstances one wishes to calculate the aggregate impact of many possible yes/no type events. For example, it is often important to answer questions such as "What is the loss amount that will not be exceeded in 95% of cases?" Simulation is usually required to answer such questions. In this model, the "yes/no" events are modeled using Binomial distributions. The results profile shows a multi-peaked distribution, which is typical when there are discrete-type inputs.
The purpose of this model is to see how many hospitals are required to accommodate all patients in various scenarios. The model assumes that patients are assigned to hospitals in a particular order: hospital 1, then hospital 2, and so on. There is uncertainty in the numbers of available beds at the hospitals.
This example illustrates a general multiserver queueing system. Customers arrive at random times. If at least one of the servers is idle, an arrival goes directly into service. But if all servers are busy, the arrival joins the end of a queue, from which customers are served in first-come-first-served order.
This example adapts the general multiserver queueing system to a city's ambulance service. The ambulances are the "servers" and a customer "arrival"; corresponds to a call for an ambulance from some location. If at least one ambulance is not currently busy, it responds immediately to the call for service. Otherwise this call waits, in first-come-first served order, for the next available ambulance. In addition to the usual queueing outputs, the model keeps track of waiting and ambulance costs per day. The number of ambulances can be increased or decreased to explore the trade-off between ambulances and waiting time.
A set of basic @RISK models for modeling events:
1. Any (or all) events could occur.
2. A version illustrating the RiskMakeInput function for use in sensitivity analysis.
3. A version where dependent events can occur (or not occur) sequentially in time.
A model that illustrates one possible generic approach for generating demands for substitute products.
An inventory ordering model that illustrates how the Time Series Define feature can be used to generate demands.
A model which illustrates how the bullwhip effect can occur with a single retailer and supplier and a second model that illustrates the bullwhip effect when there are multiple tiers of suppliers.
