| WellSpring Optimizer and Options |  |
WellSpring has a number of options for optimizing your production network. These options affect calculation scope and performance. To change the optimization, set the "Calculation Solver Type" at the Project level (Edit General data from the Data Tab - See General Data / Project Data settings).
No Contstraints
This option turns the solver off and ignores any system constraints. Wells are allowed to produce to their full capacity throughout the network.
Full Optimization
The full optimization is the main solver, and is the default.
This solver builds a large slightly Non-Linear Programming model of the network and then proceeds to optimize the total hydrocarbon production. The model that is built takes into account all of the constraints in the network and the points in the network where production can flow in multiple paths and finds the overall best solution given the constraints.
This solver will find the TRUE optimum solution to the network in the face and any topography, any number of products, and any number of constraints.
All 18 products are modeled at the same time, taking into account the constraints entered at facilities and pipelines, the topography of the network is allowed to change at each time step, and the compositional flow of each well can be unique.
In a constrained network with an infinite number of possible solutions to meet the constraints entered, the selected solution can appear rather random at the individual well level. Wells that are on production one month are given a slight edge in the model to continue production.
This model also has a slight bias against non hydrocarbon production (such as water), so that in a case where two wells can produce to a gas constrained facility, it will select the low water cut well first, for instance.
Global Sover in Full Optimization
By default, the Full Optimization uses the "Global Solver". This solver is a robust solver that improves the solution for complex (split flow) networks. As a more robust solver, it also increases the calculation time. If you are looking to improve the calculation time, you can test your project with the Global Solver turned off. The solution should still be reasonable, and the calculation time much quicker.
Full Opt Equal Constraint by Pad
This solver is based on the above Full Optimization Solver, but all constrained wells that flow to the same target facility will be cut back equally. This produces a non-optimal solution but can meet certain business constraints. The results from this solver will look more intuitive, and better behaved than the Full Optimization as production is not radically changing month to month.
If your wells flow to a lot of individual pads with a few wells each, you may still see rapid production switching by facility.
Well Priority Production
This solver uses a much more basic algorithm to assign the production to the network. To start all of the possible paths from well to the Scenario are calculated. A well that can pass through two different path’s when it encounters a split point generates a path for both directions.
Each well is assigned a priority STARTING AT THE WELL that represents each pipeline between the well and the scenario.
Each segment generates a number equal to (10 – Pipeline Priority). So a well that passes through 3 pipelines , all of which are set to priority 10 (the default) will get a priority of 0.0.0.
If the pipeline directly leaving the well was changed to priority 4 then this priority number would become 6.0.0.
Wells with the same priority are ranked in order that they were created (essentially randomly, although this order is consistant between runs).
All of the paths are then ranked against each other from highest to lowest (6.0.0 goes before 0.0.0).
Well production is then layered into that path for each well, until all of the wells production has been assigned to the network. If there is not room through any of the paths then the well is curtailed to meet the entered Constraints of the network.
This is done for each product in the model. This can mean that although there is room for the gas, the water production may then limit the well further at a later point of the calculation. The algorithm does not then try and re assign the missing gas production and mathematically sub optimal answers are possible when dealing with multiple product constraints.
Reconvergent split flow (where two paths separate then come back together later in the network, mixed with constraints at those nodes is also a situation that can have sub optimal answers and is not recommended for this solver.
Other Options for the Path Priority Solvers include:
Pre Pass Flow Percentage: This is a percentage of the production from all wells that is attempted to be flowed through the network as a first pass. Then enables all wells to have a higher chance to flow some portion of thier potential before the network is completely full. Setting this to either 0% or 100% would have no effect on the network.
Exclude PreFlow: Paths that flow through a pipeline with this check box checked will not flowed during the pre pass phase.
First Cut Priority: Paths that flow through a pipeline with this check box checked are flowed first. Before the pre pass , before the main pass. This is the easiest way to ensure particular wells are flowed without interuption reguardless of their priority settings, especially in the sales point priority model where the priorities are set from the sales point upstream.
Maximum Well Path Priority Production
This solver is the same as the Well Priority Solver, except tha the priorities are assigned a priority based on the highest priority pipeline at any point between the well and the scenario.
Wells with 3 pipelines with priorities 6,0,0 or 0,6,0 or 0,0,6 all would recieve the same priority.
Sales Point Priority Production
This solver is the same as the Well Priority Solver, except tha the priorities are assigned from the final nodes of the network upstream towards the well.
This has the impact of defining the priorities more of a pull rather than a push model in the network, letting you choose which of the incoming pipelines to prioritize at a given node.
Deliverability Balanced Allocation
This solver uses an algorithmic solution to allocate production based on prioritization of pipeline connections. This solution works great where you have an idea of the individual well/pipeline priorities in your network.
At each point in the network, the available constrained production room is allocated based on a simple weighting formula. The weighting factors are a function of the inflowing pipeline priorities. The weighting factor is equal to 10 minus the pipeline priority. A priority 2 pipeline has a weighting of 8, while a priority 10 has a weighting of zero. The allocation for each pipeline is equal to the weighting factor for that pipeline divided by the sum of all the weighting factors.
After each inflowing node has had its weighted chance to provide production, any unused allocation is repeatedly offered to any remaining inflows that have unused deliverability. This weighting is done at each individual node throughout the network starting with the most downstream nodes.
For example, if I have a priority 1 pipeline flowing into a facility with a priority 2 pipeline, then they will get allocated 53%: (10-1) / ((10-1) + (10-2)) and 47%: (10-2) / ((10-1) + (10-2)) respectively.