Abstract Gas lift is a common, widely used method of oil production. Efficient design requires knowing the performance of each component of the system. Theoretical analysis supported by experiments is needed to determine a practical model for gas-lift valve performance.
Two sets of experiments are described:to obtain standard flow performance data, and
to investigate the influence of temperature distributions on the bellows temperature of a nitrogen-charged gas-lift valve.
The new, empirically based calculation procedure presented is superior to existing procedures by:being independent of whether the constant injection pressure or constant production pressure method is used,
reducing the number of empirical model parameters to five,
addressing all three gas-lift valve flow regimes: (orifice, transition and throttling), and
predicting gas-lift valve performance more accurately than existing models.
Introduction For gas-lift operations, reliability and efficiency depend on accurately predicting the performance of the gas-lift valve. Valve performance is defined as the standard volumetric flow rate of gas through the valve as a function of the injection pressure and production pressure for a given port size, and bellows set pressure. Temperature sensitivity, corrosion and erosion resistance, and vibration suppression also affect valve performance.
The degree of temperature sensitivity has been ignored in earlier models under the assumption that the bellows temperature equals reservoir temperature. The large drop in temperature on the production end of the valve where expanding gas exits is familiar to anyone who tests valves. This indicates the lack of a single well defined equilibrium temperature in the vicinity of the valve. An experiment is reported here that assesses the sensitivity of valve performance to this temperature distribution.
Furthermore, experimental studies have identified three flow regimes in the valve with transition flow separating orifice flow from throttling flow. Existing models predict performance for orifice and throttling flow, but not for transition. They also apply subjective supplementary criteria to determine the flow regime and then select a distinct model for that regime. The new model described here applies to all three regimes and objectively predicts which flow regime applies under the design conditions. Finally, this procedure illustrates the sensitivity of the gas-lift valve to temperature.
P. 305^