San Diego Gas & Electric completed a study in April, 1986 to determine an appropriate level of peak day gas demand (and the corresponding weather condition) to be used as a criterion for planning capacity additions to the gas system. day gas demand are described in terms of their frequency of occurrence using the term recurrence interval (RI). For example, planning with a 50 year RI would require installing ample system capacity to meet peak day gas demand that would occur on average once every 50 years (22 probability of occurrence per year). These reference levels of peak The recurrence interval study evaluated seven different recurrence intervals ranging from a 2 year RI to a 100 year RI in terms of their total cost to firm gas customers. Total costs included: system facility costs (the carrying costs and O&M expenses on investments in gas system facilities), tangible outage costs (costs to restore gas service) and intangible outage costs (value of lost gas service to customers during an outage). Incorporated into the probabilistic (decision analysis) framework used in the study were uncertainties in gas system availability, demand growth and weather conditions for their determination of potential gas outages and resultant outage costs. Base case results showed total costs (expected present value over a 20 year planning horizon) were minimized between a 15 year RI and a 50 year RI, with the 35 year RI resulting in the lowest total cost. Planning for recurrence intervals greater than the 50 year RI resulted in higher total costs due to excessive facility costs. Alternatively, planning for recurrence intervals lower than a 15 year RI reduced facility costs, but at the expense of very high outage costs and higher total costs overall. An important result of numerous sensitivity cases showed the costs of over-planning the gas system (planning for too high a recurrence interval) are less severe (less risky) than are the costs of underplanning the gas system (planning for too low a recurrence interval). Based on these study results a 35 year RI for use as a criterion for planning capacity additions to the gas system was recommended to and approved by senior management.
A continual challenge to system planners and managers is making decisions to install additional gas system facilities that increase system capacity. providing service to present and future gas customers. ties in weather, the number and usage behavior of gas customers, and equipment availability; the demand on the system and system capacity is not exactly known at the time when capacity expansion decisions are made. Thus, with any gas system that could be installed, situations may exist wherein demand exceeds the throughput capability of the system.
