Towards Risk-Based Hydrate Management Strategies in Subsea Oil and Gas Flowlines Young Persons’ Lecture Competition IOM3 Western Australia Vincent Lim , PhD Candidate The University of Western Australia 25 th May 2018
Motivation: Increased Energy Demand in Future • Global energy demand is predicted to almost double in 2040 compared to 2000 • Gas contributes a quarter of the total energy production 2 Source: International Agency, 2018
Liquified Natural Gas: Alternative to Use of Diesel in Local Mining Industry Liquified Natural Gas (LNG) *1 • Predominantly methane and some ethane • Joint venture by local petroleum promote LNG over diesel in mining industries • Cleanest-burning fossil fuel • Less environmental hazard *2 • Natural gas resources are plentiful in Western Australia • Q: What are the challenges for gas production in pipelines? Hydrates! 1* Hussein, A., Oil Industry Insight , 2018 3 2* Government of WA, Department of Mines and Petroleum, North West Shelf Oil & Gas, 2015
What are Gas Hydrates? Gas hydrates • Ice-like solid compounds • Small gas molecules (e.g. CH 4 ) trapped within hydrogen bonded water cage • Stable at high pressure and low temperature Water Experiment: Hydrate crystal formed in Autoclave Methane molecules molecule $ 1M lost per day if pipeline plugging occurs! Pipeline: Hydrate plug (Image from Petrobras) Hydrate stability region Hydrate- free region 4
Natural Gas Hydrate Formation is Stochastic (aka random) Q: Will water be immediately converted to ice at 0 ° C? Field test: Werner Bolley well, Hydrate stability Southern Wyoming region *1 Δ T 2 Δ T 1 Minimum subcooling Hydrate-free region Formation Pipeline Hydrate equilibrium curve • 4 measurements of hydrate formation in field • Minimum subcooling of 6.5 ˚ F ( ≈ 3.6 K) • Used in hydrate simulation software: OLGA Q: Is this 3.6 K universally applicable to oil and gas pipelines? 5 *1 Matthew et. al., Annals of the New York Academy of Sciences , 2000
HPS-ALTA is more Efficient in Generating Hydrate Formation Event High Pressure Stirred Automated Lag Time Apparatus Traditional Apparatus: • Stirred system (to remove mass transfer limitation) Autoclave (~2 K/hr) • Peltier driven cooling system (2 K/min) • Able to generate large number of experimental runs Data are not (typically on the order of 100 data points) statistically significant 0 rpm 700 rpm rotate 6
Methodology: Hydrate Formation Detection by Pressure Drop Hydrate Initial condition formation induced gas consumption Δ T Hydrate dissociation Note: 100 formation data points collected for each experimental condition 7
Hydrate Formation Depends Critically on System Shear During pipeline shutdown (0 rpm) • No pressure drop (formation) was observed in static system 0 rpm 100 rpm For 100 rpm • Probability distribution (PDF) is generated from > 100 hydrate formation events • Cumulative distribution (CDF) generated by numerically integration Mean 8
700 rpm is more representative of hydrate formation during production Increasing shear rate from 100 rpm to 700 rpm decreased mean subcooling by approximately 3 K 100 rpm 700 rpm *1 Mean Note: Low subcooling indicates “easier” formation 9 *1 Lim et. al., OTCA , 2018
Hydrate Prevention by Chemical Injection: KHIs is More Economical than THIs Traditional solution: Into the millennium: Thermodynamic Hydrate Inhibitors (THIs) Kinetic Hydrate Inhibitors (KHIs) Key Points Key Points • Methanol / glycol • Delay formation • Completely prevent hydrate formation • Typically 0.5 to 2 wt% added, reduce • Uneconomical if high dosage logistic costs and OPEX (typically 50 wt%) is required • Mechanism of KHIs to delay hydrate formation is unclear • Will you risk using KHI? 10
KHI Suppress Nucleation, Less Stochastic Distribution Obtained Key Points • Mean subcooling obtained with 1 wt% KHI increased 3.3 K • Standard deviation of distribution with KHI decreased by 3 to 5 times *1 Conventional apparatus Increase KHI concentration With large formation datasets, we can fit experimental datasets to hydrate formation theory. Modelling cannot be done with data collected with conventional apparatus with repeatability of 5 points! 11 *1 May et. al., Langmuir , 34, 10, 3186-3196 2018
Conclusion and Way Forward Summary 1) Increased risk of hydrate formation as natural gas production moves towards deeper subsea region 2) HPS-ALTA is efficient in generating statistically significant datasets to study hydrate formation 3) Detailed characterization of KHI performance 4) Hydrate formation distribution data can be fitted with hydrate nucleation theory Way Forward • Rank performance of different inhibitors • Perform modelling work on hydrate formation data • Promote hydrate formation model to industry (e.g. OLGA) 12
ACKNOWLEDGEMENTS Peter Paul Mike Zachary Eric Metaxas Stanwix Johns Aman May THANK YOU QUESTIONS?
Comparisons of Conventional Apparatus to Screen Hydrate Formation Autoclave: High pressure stirred cell with impeller *1 • Direct P & T measurements • Visual observation • Cooling rate: ~1 K/hr • ~10 hours per run • Formation data are not statistically significant *2 High Pressure Automated Lag Time Apparatus (HP- ALTA): Mini high pressure cell • Sudden decrease in transmitted light intensity when solids form • Large range of cooling rate: up to 5 K/min • Can generate lots of formation data • Only work in static system • Only work for transparent sample *1 High pressure autoclave owned by UWA 14 *2 Maeda et. al., Review of Scientific Instruments , 82, 065109, 2011
Methodology: Hydrate Formation Detection by Pressure Drop Experimental Methodology a) Constant cooling rate Formation T b) Hydrate forms indicated by pressure drop Subcooling c) Hold temperature at T low for 5 minutes d) Heat the cells to regeneration temperature Equilibrium T a d Python automation c b 15
Statistical Data for KHI Experiments 16
Recommend
More recommend
