A Probabilistic Model To Incorporate The Impacts Of Floods Into Production Forecasts And Development Planning For The Cooper Basin Gas Fields
Petroleum Engineering Honours Degree, 2011
University of Adelaide
Located in Central Australia, the Cooper Basin is the largest river basin in Queensland. It is a poorly understood hydrological system due to its abundant anastomosing channels and low-lying floodplain, renowned for extreme and highly variable weather conditions. Santos has been producing hydrocarbons from the region for the past forty years, of which twenty have experienced varying magnitudes of flooding, hindering the ability to maintain production forecasts, transport equipment, undertake exploration and development, and manage logistics. The project aimed to assess the relationship between flood frequency, severity and subsequent impact on Santos’ Cooper Basin operations.
To date, methods to determine the frequency and severity of flooding in the Cooper Creek river system have been tedious and slow to develop due to limited and sparse data within the expansive study area. Key literature by Knighton and Nanson (1999, 2001), Puckridge et al. (1999) and Amos and Jaeregui (2010) have proposed several different methods to analyse flood events in arid zones of Australia. This project collaborated and further developed existing approaches, assessing the frequency and severity of floods based on key indicators. These include rainfall data, stream flow data and peak river height and the climate-based indicator, Southern Oscillation Index (SOI). The frequency of flood events in the Cooper Creek was analysed by producing average recurrence interval (ARI) and annual exceedance probability (AEP) plots, based on annual maximum stream flow. A flood classification system developed by Puckridge et al. (1999) was used to determine the severity of flood events at Cullyamurra, neighboring Santos’ Moomba operations. In addition, satellite images were obtained to provide further insight into flood magnitude and duration. The relationship between SOI and rainfall was investigated to assess the ability of SOI in predicting significant rainfall events.
The impact of flooding on Santos’ operations was investigated and summarized using articles and reports sourced both online and from the Santos Library and Santos’ intranet. Key findings were presented with respect to flood severity classification in order to effectively illustrate the relationship between impact and severity. Although descriptive and qualitative in nature, these findings provide a comprehensive reference for understanding the potential impact of future floods on operations.
Several key results were identified through the analysis undertaken for this project. Puckridge et al. (1999) provided the best flood classification methodology based on producing a frequency histogram with a normal distribution. The ARI and AEP plots enabled extrapolation beyond the known data to determine the expected magnitude of the 1 in 100 year flood event. It was found that consecutively high SOI, found in the literature to represent periods of climatic wetness, does not indicate the occurrence of a flood event, whilst this does not hold true for the opposite. Furthermore, the flood events of 1974, 1989 and 2010 were found to have the greatest impact on Santos’ operations.