Opitimized Oil Production Plan
GOMAN – GMANOPT/SIM
Economically optimizes development of oil and gas complexes – reservoirs producing through parallel gathering and processing facilities to one or more offtake points. Water & solution gas (AG) are separated from oil, which is then stabilized and graded for sale. Non-associated (NAG) gas passes through a separator. AG, NAG and stabilizer & separator bottoms flow into gas plants, which output acid gas (→ sulfur plant), sales gas & natural gas liquids (NGL). The latter is split into components in an NGL plant. Produced water is injected along with makeup water into water flooded reservoirs. Aramco, Kuwait Oil, Santos
GOMAN’s domain is a collection of reservoirs whose reserves are to be produced and sold at a targeted rate over time. Reserves may be crude oil or non-asssociated gas (NAG), or a mixture thereof. GOMAN sets up an optimizing model of the operation spanning a designated planning horizon – typically 20 – 25 years. The model’s solution is an optimal sequence of capacity additions and expansions that meets the targeted rate for as long as can be done economically with the available reserves. With insufficient reserves GOMAN maximizes economic rate – minimizes shortfall – during the latter years of the horizon.
Although many of the equations that describe the behavior of the production-processing complex are nonlinear, the optimizing model is a form of linear programming (LP). Linearized equations approximate the nonlinearities within the LP. However, the nonlinear full scope is handled by trial-and-error. A new LP that more closely approximates the current state of the system is generated from the values – flows, expansions – of the previous LP solution.
Still another LP is generated from this new solution – and so on until convergence, at which point the linear equations in the LP approximate closely the nonlinear equations. The LP’s objective equation sums the DCF of all revenues, investments & operating costs during the planning horizon & its solution maximizes this sum.
GOMAN traces the capacity of each reservoir with a rate-cum curve. For oil reservoirs these are derived by engineers using simulators or other reservoir analysis tools. Input for each oil reservoir includes CAPEX ($/well), OPEX ($/well/yr), initial well capacity (Bbl/D), and decline fraction as a function of cumulative. Besides wells, capacity-increasing options are submerged pumps, gas lift and water injection. Using this input GOMAN sets up development over time of each reservoir to supply the take called for by the optimal solution. An overview of the process is – given a collection of reservoirs GOMAN weighs cost vs deliverability to find the most economic time at which to produce each.
For each NAG reservoir, GMANOPT/SIM determines the rate-cum curve along with the optimal schedule of wells and compressors. As for oil, GOMAN slots the take from each NAG reservoir into the schedule at the time most advantageous for overall economic performance.
For an oil reservoir, production rate of associated gas (AG) and water are calculated using curves of GOR & WOR vs time. Water & AG are removed from oil in field separators. Each reservoir’s oil, after stabilization is rated for quality (%S, oAPI) and blended into a grade for shipment. Price of each grade depends upon quality. Water is injected along with makeup water into producing oil reservoirs, or into disposal wells. Separators for NAG reservoirs remove produced water and an NGL bottoms stream.
AG & NAG (plus separator bottoms) flow into gas plants which remove acid gas (H2S, CO2) and separate the remainder into sales gas & NGL. Elemental sulfur is recovered from acid gas in sulfur plants. NGL is separated into sales components (C2, C3, C4, C5+) in NGL plants. Sales gas flows directly to market. GOMAN monitors capacity of all surface facilities. Using CAPEX & OPEX of each, the optimal sequence of expansions – separators, gas plants, NGL plants, water supply, etc – is determined
GOMAN generates an array of output. Graphs show every flow rate vs time – oil, gas, water for reservoirs, areas, etc. Shown above is a stacked plot of production rate from all oil reservoirs during a planning horizon. Composition of gas steams is tracked. This example shows annual flow of moles of each component into an NGL plant.
The tabular portion of GOMAN’s output is an extensive set of files. Oneset displays complete physical performance including rate of production of oil, gas, water – thruput rate vs capacity of all surface units – output rate of crude oil, sales gas, NGL components. sulfur, etc. Another set displays complete economic performance, including CAPEX & OPEX of each production and processing unit. Other setsgive the distribution of sales revenue and performance measures & ratios – DCF breakdown, DCF$/Bbl, Cost$/MMCF, etc. Another setpasses complete results of a run into PEEP.
GMANOPT/SIM’s domain is one or more NAG reservoirs producing thru a network to a single offtake. Here is a graph of a three-reservoir example (Res1, Res2, Res3) taken from DAT, SIM’s input data module. DAT writes an input run file for a simulation.
The NAG simulator – SIM – divides the formation into (up to 100, usually smaller) grid blocks (called regions). Here, Res3 has one region – a “tank-type” model. With “normal” NAG (connate water and dry gas) viscosity is low and pressure gradient in the formation is small. Then, a tank-type model – with pore volume (Vp) split between the nw wells to compute drainage area and GIP/well – often gives a satisfactory approximation.
In Res1 (3 regions) and Res2 (2 regions) NAG flows between regions when pressures become unequal. Blue fill – Res1 & Res2 – signals an aquifer in all regions – modeled with Fetkovich’s equations (Dake, pp 325-333). White fill denotes no aquifer in Res3.
Each reservoir has 1 – 15 producing layers. In a single-layer reservoir wells may be vertical or horizontal; with more layers only vertical wells are allowed. Vertical wells penetrate all layers, but a layer produces only after being perforated – and until plugged. For each well a vector of perforation/plug dates is input in DAT along with each region’s drilling schedule giving number and date of well additions.
The red lines connect wellheads to pipelines. Circles (nodes) denote junctions. The icons at Nodes 2 & 250 represent compressors. Any node can have a compressor. A compressor is active if horsepower (HP) is input via its schedule in DAT.
On each time step SIM solves a “nodal analysis” of the system. The equations consider all pressures and flows rates – in the formation, wellbore and surface facilities – simultaneously. Of SIM’s various run options GMAN.OPT uses ‘station rate vs time’. Here, SIM draws NAG from any or all reservoirs to meet the requested rate profile at Node 250. The standalone SIM delivered as part of the GMANOPT/SIM package (along with GRIDDER, DAT & TFR) will make simulations with all run options. Maraco is now supporting changes to GMANOPT/SIM but not to standalone SIM. GasPal is the standalone NAG simulator that Maraco now supports.
GMANOPT’s domain is the same as SIM’s, since it contains SIM as a .DLL. OPT determines the optimal development schedule (wells, compressors) for the group of NAG reservoirs to meet a targeted offtake rate for the planning horizon.
Shown here is a portion of the user interface during an OPT run. The top graph depicts Project structure. The red icons at the top refer to the two compressors shown in DAT’s window. Platform is simply the project heading, beneath which Res1, Res2 & Res3 are listed with their regional breakdown.
The lower portion is OPT’s input/output panel. When OPT starts, it reads pertinent data from the SIM file and writes an OPT input file with default values. A click on Edit Data brings up panels to edit the latter. OPT has both Graphic & Tabular output. The latter gives the investment tests and the resulting optimal schedule plus an economic summary. A click on To Sarah yields a text file with the optimal production & investment schedule for input into GOMAN. A click on Export Schedules yields a text file containing the optimal schedule. DAT reads this file & writes a SIM file that SIM reads, allowing the user to conveniently make runs to examine & tailor the schedule as desired.
OPT fills the schedule in an annual sequence starting in Year 1 of the horizon. The computation procedure is the same in each planning year. At start of a year (or an iterative loop within a year) OPT first determines production rate with the wells now (or initially in Year 1) in place. If the target is met, calculations move to the following year.
If the target is not met, OPT puts one well into each region in turn, and calls SIM to calculate the offtake rate vs time. After calculating the incremental NAG rate, OPT determines the well’s (incremental) DCF, PWI (Present Worth Investment) and PVR (DCF / PWI) using CAPEX & OPEX for the region’s well and the model’s NAG price vector. (For offshore wells costs includes an allocated portion of platform costs.) OPT also calculates the increment in the model’s production rate, ∆q. When a well in each region has been evaluated, OPT ranks them by a preference criteria, Xw. When all region’s have been evaluated, one or more wells are added in the region with the largest Xw and PVR > PVRmin & ∆q > ∆qmin. The number of wells added varies with PVR.
After a well addition, OPT loops back to start another determination. This iterative sequence continues as long as wells that satisfy the decision criteria are found. If a year is reached in which all wells are uneconomic or only marginally so, or if ∆q < ∆qmin for all economic wells, OPT checks to see if adding HP is an economic alternative – provided compression is allowed. Compressor evaluations parallel the procedure for wells. If before the end of the planning horizon a year is reached in which the target is not met but all wells and compressors fail the selection test, OPT terminates the run.
Graphic output from OPT displays the investment schedule and production rate. The Project Production Profile from an evaluation of the Res1-Res2-Res3 model shows requested target (dmd) rate, total capacity (cap) and NAG produced (prd). With a 20-year planning horizon, dmd could be supplied economically only through Year 11. The top line labeled C1 lists the (5 & 6) increments (user setting of 300) of HP added – a total of 3300 HP. The second graph shows production rate and drilling schedule for Res1.
PRICE: $70000 + 25,000 contracted Annual Maintenance
(Software is for unlimited ownership of purchased version. Annual Maintenance is for improvements, access to latest versions, fixing any bugs,
plus 24 hour help on a problem encountered)
Discounts Given on Two or More Software Orders
* Note – None of these prices include your local agent or local
Contact Me If You Need Further Information Or Wish To Place An Order: