The demand of natural gas follows seasonal patterns with a peak in winter and trough in summer. But the production rate of natural gas is almost constant and generally not agile enough to meet peak demand in winter. A common business practice is to store natural gas at storage facilities in preparation of demand peaks.
Alongside demand, the price of natural gas follows these same seasonal patterns with a peak in winter and trough in summer. To maximize profitability, a nat gas trader will exploit the price differential using spot price and forward market price information. Likewise, the storage facility operator (either the owner or lessee) will aim to inject gas into the storage facility at low prices and withdraw gas from the storage facility at high prices, to maximize the benefit of the price differential. However, the storage capacity at a facility is not unlimited and there are technical considerations that influence how much natural gas can be stored in inventory.
- A minimum level of natural gas at a storage facility is required to maintain pressure.
- Natural gas storage facilities are limited to a maximum quantity for which they were designed.
- There are limitations on injection rates and withdrawal rates based on levels of gas in the facility.
Determining the Influencing Factors
Companies can use ETRM software to manage the lifecycle of natural gas, including scheduling. But to maximize profitability, storage facility operators need to determine the optimal natural gas inventory and schedule of injection and withdrawal during the storage contract to maximize the benefit of price differentials. This requires advanced analytics as the resulting schedule will depend on several factors including:
- spot price
- future prices for upcoming months between the contract start and end dates
- fuel loss during injection and withdrawal
- cost of injection and withdrawal
- interest rates
- operational constraints
This is an ideal problem to solve with linear programming.
Optimizing the Schedule for Natural Gas
At each decision point, we need to answer the following questions:
- Should we inject gas into the storage facility? If so, how much?
- Should we withdraw gas from the storage facility? If so, how much?
- Should we do nothing?
By supplying the necessary inputs into a model, we can answer these questions and determine the optimal schedule. The inputs to the model include: details about the storage contract, injection withdrawal rate limits, injection withdrawal costs, minimum natural gas inventory level requirements, and forward prices. The result from the model provides a schedule as follows: when prices are low, the model recommends injection and when prices are high, the model recommends withdrawal.
Asking the "What-if" Questions
From this initial scenario, we can modify the input data to answer those "what-if" questions, such as:
- What if there's a price spike in a given month?
The model recommends a withdrawal in that month (instead of a previously recommended injection) with a following injection in a future price advantageous month. - What if the injection and withdrawal rate limits are increased?
The model recommends higher injection and withdrawal at certain points. - What if there's a rise in interest rates?
The model recommends early withdrawal and discourages early injection. - What if there's a change in the cost of injections and withdrawals?
The model recommends any changes to injections and withdrawals including changing quantities. - What if losses due to injections and withdrawals increase?
The model recommends where changes are needed in the schedule.
Analytics solutions designed for commodities markets have been developed that address these very specific types of questions. (Learn the benefits to be gained with commodity-specific analytics solutions in "The Benefits of Analytic Solutions Focused on Commodities.") Eka's Commodity Analytics Cloud has been built with the specific needs of commodity companies in mind. The platform delivers industry-specific apps that cover P&L, procurement, risk, and supply chain.
See more details of Eka's model for natural gas storage optimization in the white paper, "Using Linear Programming to Optimize Natural Gas Inventory."