(1. INTRODUCTION
The advance of technology in recent times has given another perspective to industrial processes, displacing arduous human labor with many opportunities for the supervision of computerized processes (Acharya et al., 2017; Chakraborty et al., 2020; Viraktamath et al., 2020). This has been achieved through the automation of complete systems and processes, where the parameters to be followed by the equipment are programmed to carry out a specific job or the monitoring of the process to maintain its proper functioning (Kiangala & Wang, 2019; Tomar & Kumar, 2020; Tomar et al., 2023).
Automation is a set of technologies that allow machines and systems to work without the continuous intervention of a human operator, and thus achieve superior performance compared to manual operations (Abdu & Taleb, 2014; Ibrahim et al., 2018; Inzunza et al., 2020; Li et al., 2021; Mendoza, 2016). In this context, today's industrial processes are oriented towards achieving a continuous increase in operational reliability, and they have identified opportunities for improvements related to the reliability of their equipment and automation systems (Hussein et al., 2022; Kiangala & Wang, 2019; Nadgauda & Muthukumaraswamy, 2019; Tomar et al., 2023).
Currently, the trend in the field of automation and process control aims to improve operations carried out by various manufacturing companies, batch production, continuous production, among others, by replacing existing manual systems with fully self-controlled and reconfigurable processes (Alphonsus & Abdullah, 2016; Kiangala & Wang, 2019; Yu & Xue, 2022). Specifically, oil companies characterized by continuous production need permanent monitoring and control in most of their operations (Díaz, 2019; Pino, 2013), to obtain benefits by optimizing the said production and avoiding the occurrence of incidents and accidents within facilities that can cause material and human damage (Ronceros et al., 2023; Tomar et al., 2022).
In this context, industrialization has boomed in the main areas of production, control, and maintenance of the oil and gas industry, where it depends on the use of information technologies as a means to achieve an efficient level of operation (Estrada et al., 2017; Fletcher et al., 2019; Zawawi & El-Sayed, 2012). In this sense, the incorporation of automation into oil and gas industry production processes has generated an increase in the levels of safety of production processes and operators (Cotrina, 2018; Quispe, 2019; Villalba, 2019), which optimizes production time by reducing user interference with the general process, leading to increased productivity, better quality, efficiency, and reduced labor costs, as well as human errors in industries (Corvo, 2018; García et al., 2018; Lashin, 2014).
Regarding the benefits of automation in oil production processes, the results obtained by Ilyushin (2022), who implemented an automated process control system to prevent the formation of kerosene, asphalt, and resin deposits in oil wells with low performance, optimize the production of paraffinic oil, reduce energy costs, and improve operational efficiency, stand out. Similarly, Ronceros et al. (2023) developed a supervision and control system for operational variables (flow, temperature, and pressure) in a cluster of a high-pressure gas injection plant, which allowed optimizing gas flow regulation, obtaining a reduction in errors in the opening of control valves, going from a range of 2-5 % with manual control to more accurate values with the automated system.
Like Wu et al. (2022), who developed an automatic control system to optimize phase separation (water, oil, and gas) in horizontal three-phase separators, thereby improving separation efficiency and reducing the water content in crude oil. The proposal increased the efficiency of dehydration to 95.71 %, which is a significant improvement in the flow of water and a reduction in the water content in the oil. In the same way, Safarova et al. (2023) implemented a multiconnected control system to regulate parameters such as temperature, pressure, and flow at various stages of the coking process. This proposal improves the stability of operating parameters, minimizing external disturbances and fluctuations in the characteristics of the processed crude oil. Ronceros et al. (2024) designed a gas flow control system for oil wells through artificial gas lift (Gas Lift), allowing the increase in the productivity of the wells through optimized gas flow control. A networked data acquisition system (Net-DAS) and the GALBA SCADA system were used through a control logic with four states: local, remote, automatic, and optimized remote.
In this context, the automation of an oil outlet manifold is presented as a strategic solution to optimize the distribution and control of crude oil flow to different processing or storage units. The implementation of advanced control systems in the outlet manifolds allows for a more efficient management of the operating variables, ensuring a balanced and adaptive distribution of oil according to the process demands and operating conditions.
The Crude Storage and Transportation Center Patio de Tanques Travieso (PTT), located in the north of the State of Monagas, Venezuela, receives crude oil production from the flow stations located in the Punta de Mata, El Tejero and Jusepin operations area of the Furrial District of Petróleos de Venezuela S.A. (PDVSA). At this center, two types of segregation are handled, medium crude and Santa Bárbara light crude. The operational objective of the crude oil transmission system in PTT is to provide sufficient pressure to send it to the Jose storage and shipping terminal, the Guaraguao storage and shipping terminal (TAEG), and the storage center and transportation of crude Anaco. PTT has a set of pipes where the discharges from the main pumps converge and then join the pipeline that transports the crude oil to its final destination. This set of pipes is called the output manifold, and its objective is to achieve efficient pumping of crude oil. This multiple is made up of two 26 and 30-inch pipelines with which medium crude is handled. The light crude is pumped through a 16-inch pipeline located parallel to the outlet manifold.
In this scenario, the Patio de Tanques Travieso Crude Oil Storage and Transportation Center (PTT), located in the north of Monagas state, Venezuela, receives crude oil production from the flow stations, located in the area of operations of Punta de Mata, El Tejero, and Jusepin of the Furrial District. At this center, two types of segregation are handled, medium crude and Santa Bárbara light crude. The operational objective of the crude oil transmission system in the PTT is to provide sufficient pressure to send it to the Jose storage and shipping terminal, the Guaraguao storage and shipping terminal, and the storage center and transportation of crude Anaco. The PTT has a set of pipes where the discharges from the main pumps converge and then join the pipeline that transports the crude oil to its final destination. This multiple is made up of two 26- and 30-inch pipelines with which medium crude is handled. The light crude is pumped through a 16-inch pipeline located parallel to the outlet manifold.
The control of the discharge pressure in the main pumps and pipelines, a process that regulates the flow of transported crude, is carried out manually and semiautomatically, for which the operator must be monitoring the information transmitted in the control room and making the necessary adjustments to maintain the valve management in the corresponding parameters, and thus control the flow and pressure to achieve optimal delivery of crude oil. In this regard, in view of the great importance of pumping crude from this station to the Jose Crude Storage and Shipping Terminals, the Guaraguao Crude Storage and Shipping Terminals, and the Anaco Crude Storage Center and Transportation, the present investigation is oriented to the design of a proposal for a control system and a technological architecture that allows operations to be carried out automatically avoiding the presence of operators in the field.
This research is structured as follows: Section II describes the methodology used to approach the research and develop the proposed monitoring and control system. Section III presents the development of the output manifold monitoring and control system and the corresponding discussion of results. Section IV concludes the work and highlights future work.
2. METHODS
The development of the research was based on the methodology of the Capital Investment Project Management Guides (PDVSA, 2021). The MGCIP is structured in five phases: visualization, conceptualization, definition, implementation, and operation. The project only covered the first three phases since it is an automation proposal. Each of the phases is described below:
Phase I: Visualization. In this first phase, the required information was gathered in order to understand the operating philosophy and identify infrastructure requirements. Among the activities carried out in this phase are the following: Identification of PTT processes, identification of the current control philosophy of the PTT output manifold, and identification of the current control philosophy of the main PTT pumps.
Phase II: Conceptualization. In this second phase, the minimum functional and integration requirements for the control system were formalized namely a modular, scalable architecture, and seamless interoperability with the ALBA Guardian SCADA. A weighted multi-criteria evaluation matrix was constructed to operationalize the selection criteria and their relative importance. Three candidate platforms were benchmarked against vendor technical documentation and corroborating operational evidence from PDVSA. The weighted analysis identified the highest-scoring option, which was selected as the target platform for subsequent design and integration.
Phase III: Definition. The objective of this phase was to develop the detailed engineering of the proposed automation of the Outlet Manifold of the Storage and Transportation Center of the Tank Yard. The activities carried out were as follows: Establishment of the control philosophy that will cover the operational needs of the output manifold, integration and design of the new technological architecture, design of the PLC control logic that complies with the new specifications presented, and design of the control and supervision schematic in SCADA guardian alba of the output manifold.
3. RESULTS
3.1 Stage I: Visualization
3.1.1. Identification of PTT Processes
The PTT is made up of thirteen atmospheric tanks, built with steel sheets welded together, complying with the specifications of the API 650 standard, with capacities of 97, 130, 200 and 250 thousand barrels. Four of these tanks store Mesa 30 crude from the flow stations: Muri, Musipan, Carito, and Amana Operations Center; another four crude from the Jusepin Tank Farm, and the remaining five tanks store the Santa Barbara segregation of the Tejero Operations Center, Santa Barbara, the Amana Operations Center, and the Jusepín Tank Farm. The PTT establishes a control philosophy that contains all the parameters that must be followed when filling tanks for the storage of crude oil and then for its subsequent unloading.
Pump Rooms 1 and 2
It consists of ten motor-pump units. Pump room 1 has a pumping capacity of 3 500 barrels per hour and Pump room 2 has a pumping capacity of 4 500 barrels per hour. Each pump has a fan to maintain normal temperature the motor of said pump, and a pre-lubricating pump with the function of lubricating the motor at the time of starting and stopping it; likewise. Likewise, the system has a number of associated instruments.
3.1.2 Identification of the current control philosophy of the PTT output manifold
Currently, there is no established control philosophy; this manifold was improvised as oil production increased at the different flow stations that send the crude to it (PTT). The operation of the output manifold is directly related to the main pump rooms 1 and 2, which are the main axis of the entire tank farm, since most of the actions carried out in the PTT are executed by these pumps from the filling and emptying of the tanks, until sufficient pressure is generated to send the crude toward the output manifold. Based on this, it is necessary to take into account the control philosophy of the main pumps.
3.1.3 Identification of the current control philosophy of the main PTT pumps
The start of the main pumps in the PTT can be done manually locally, manually remotely, and semi-automatically, with the latter two modes available on the console. Room No-1 when giving the automatic sequence command to start the pump for medium and light crude, the fan and the main electric motor turn on simultaneously, the suction and discharge valves remain open, they are not motorized, and there is no prelubrication, following a sequence similar to that described for manual mode. The system was configured according to the type of crude oil, medium (crude oil whose API gravity is between 22.3 ° and 31.1°) and light (crude oil whose API gravity is greater than 31.1°).
In the medium system, after pump activation, discharge, and pipeline pressures are controlled through a control valve installed at the exit of pump room No-1. In the light system, after the pumps, the discharge and pipeline pressures are controlled through the pump gate valves.
In Room No. 2, when the automatic start sequence for the medium-crude pump is issued, the pre-lubricator pump and the fan energize. The motorized suction valve then opens, once it reaches 100% open, the motorized discharge valve begins to open, and the main electric motor starts the pump. After three minutes, the pre-lubricator pump turns off and will re-energize only if the electric motor stops or if low lubrication pressure is detected in the motor’s lubrication system. In the medium system, after pump activation, discharge, and pipeline pressures are controlled through a control valve installed at the exit of pump room No-2.
3.2 Stage I: Conceptualization
The minimum requirements that the proposed control system must have are the following.
Modular and scalable design. It allows the system to be autonomous because at the time of a failure, it will only affect a specific element without affecting the rest of the operation of other devices. It also facilitates the expansion of basic components.
Programming and multidisciplinary control. Compliance with IEC 1131-3 programming languages for programming PLCs.
The system should facilitate its integration with the ALBA Guardian SCADA. ALBA Guardian is a software developed by PDVSA that integrates high-level functionalities to solve process monitoring and control applications. It uses a distributed architecture of modules that allows for scaling to large-scale applications. The system communicates through DF1, DNP3, Modbus (IP and Serial) and Ethernet/IP protocols, therefore a Controller (PLC) compatible with these protocols is required to guarantee communication with the supervision system, complying with PDVSA's Technological Component.
The system must have communication to and from the Ethernet network. The proposed PLC must be compatible with this protocol, in order to ensure communication with the supervisory system;
Technical support. The manufacturer must have the availability and facility to provide technical advice and training options to personnel.
Costs. The selected equipment must be economically accessible.
Identify the minimum requirements that the new control system must have. A technical study was carried out to analyze the different technologies that can meet these requirements. It is important to note that to carry out the study, the control systems currently used in the PDVSA platform were taken into account to facilitate the homogeneity of the platform, since there is a wide range of manufacturers and models on the market of control systems.
Control System 1 (ControlLogix)
It uses a common control engine with a common development environment to provide high performance in a user-friendly environment. The tight integration between the programming software, controller, and I/O modules reduces development time and lowers the cost at start-up and during normal operation. It also enables standard and safety control on the same chassis in a truly integrated system. This model has the following characteristics:
Provides high-speed, high-performance, multidiscipline application control.
Provides a fully redundant controller architecture with seamless switching and high availability.
Provides modular and scalable network communications.
Offers a range of broadcast communication and I/O options.
Provides time-synchronization capabilities for first failure, process sequencing, and other applications.
Offers some products with TÜV certification for use in SIL 2 and SIL 3 applications.
Control System 2 (Modicon Quantum)
Provides a wide range of processors and is suitable for complex processes. The power of its processors optimizes cycle times and, at the same time, integrates more and more communication, diagnostic, memory flexibility, and data storage functions. It provides a well-balanced CPU capable of excellent performance from boolean to floating-point instructions. It is ideal for complex process control and has the following characteristics:
IEC languages as standard: LD, ST, FBD, SFC, IL.
High-level multitasking system.
Memory capacity of up to 7 Mb using PCMCIA extensions.
Specially configured for process control applications with conformal coating modules, intrinsically safe I/O, and a wide catalog of associated modules.
Safety processors and I/O modules.
High-performance plug-and-play hot standby solutions with an LCD keypad for local monitoring.
Numerous built-in ports (ISB port, Ethernet TCP/IP port with Web server, Modbus Plus, and at least one Modbus serial port) are located on the front panel.
Rack connectivity to Profibus-DP.
Control System 3 (Simatic)
It allows the use of a wide range of field buses, such as AS-Interface, Profibus-DP, and Ethernet. The STEP 7 programming environment makes all industrial PLC programming languages available: STL, LDR, FCH, STEP 7-SCL, STEP 7-GRAPH, and STEP 7-HiGraph. Modular controllers can be flexible to any time through pluggable I/O, functional, and communication modules, providing custom solutions for your needs. Modular PLCs are also presented as high-availability or fail-safe systems. These have the following characteristics:
Standard RS-485 port with data transfer rate between 1.2 and 187.5 kbit/s.
Ethernet connection through a dedicated module.
Windows standard.
Figure 1 shows the control system evaluation matrix. Figure 2 indicates that Control System 1 is the one that meets the requirements, obtaining 148 points over Control System 2 with a total of 142 points and Control System 3 with 127 points. Regarding the criteria related to modular and scalable design, it obtained a score of 50; similar to Control System 2 and Control System 3. The criterion corresponding to programming and multidisciplinary control obtained a score of 25 because it complies with the standardized languages in the IEC 1131-3 standard (Ladder logic or contact diagram, instruction list, structured text, and functional block diagrams), as the other two control systems with the same number of points, respectively, for Control System 2 and Control System 3.
The criterion related to Integrability to the Guardian del Alba SCADA system obtained a score of 35; as well as Control System 2 and Control System 3. Regarding compatibility with the Ethernet protocol, Control System 1 obtained a value of 5, while Control System 2 and Control System 3 obtained an amount of 4. On the other hand, for technical support, scores of 25 points were obtained corresponding to Control System 1, followed by Control System 2 with 20 points and finally Control System 3 with 5 points.
3.3 Stage III: Definition
In this phase, once the requirements for the new control system were analyzed and the platform assigned to carry out this project was considered, a proposal was made to automate the output multiplex to optimize crude oil pumping. For this, the new control philosophy, the design, and configuration of the PLC were carried out, and in the same way the deployment of the multiple output with the Guardian of the Alba system was designed.
3.3.1 Establish the control philosophy that will cover the operational needs of the output manifold.
The start of the main pumps in the PTT can be done manually locally, manually remotely, and semiautomatically, with the latter two modes available on the console. Room No-1 when giving the automatic sequence command to start the pump for medium and light crude, the prelubricator pump is activated and immediately begins to open the motorized suction valve when it is 100 % open, it begins to open the motorized discharge valve, and immediately the main electric motor is turned on to start the pump, after several minutes the prelubricator pump is turned off, it is only activated again when the electric motor stops or when there is some low lubrication pressure in the electric motor.
In the medium system, after activation, the pumps are activated and the discharge and pipeline pressures are controlled through a control valve installed at the exit of Pump Room No-1. In the light system, after pumps, the pressures of the discharge and pipe lines are controlled through a control valve installed at the exit of pump room No-1.
In Room No-2 when the automatic sequence command is issued to start the pump for medium crude, the pre-lubricator pump and the fan are activated. The motorized suction valve then begins to open when it is 100 % open and it starts to open the motorized discharge valve. Immediately afterward the main electric motor is turned on to start the pump. After several minutes, the pre-lubricator pump is turned off; it will only activate again when the electric motor stops or when there is some low lubrication pressure in the electric motor. In the medium system, after pump activation, discharge, and pipeline pressures are controlled through a control valve installed at the exit of pump room No-2.
Initial conditions:
The start-up process for the pumping of medium and light crude to Jose and Guaraguao will begin as long as the availability conditions of the pumps are verified by the control system, according to the status of each of the instruments associated with each one of them and the permissive set. These permissives are indicated below, using the P-11 bomb as an example.
Pump in service (defined by the Operator)
Selection of the pump BP-500-A
Temperature protection system available
Vibration system available
Electric motor available
The oil pressure is good
Motorized suction valve, completely closed
Suction pressure (confirmed with the operator), not less than 2 psig
Differential pressure on the suction filter, operational
Fully closed motorized discharge valve
The high pressure at discharge is not activated
Pressure at discharge
Operation under normal conditions:
Once the control system has verified the above conditions, the start of the BP-500-A pump can be started as indicated below.
The automatic sequence of the selected pump BP-500-A is activated.
The engine prelubrication system starts and the pump cooling system fan is activated.
Once the status of the engine Prelube system and the pump cooling system fan are in place and the appropriate Prelube pressure (23 psi) is confirmed, the motorized suction valve starts to open.
Once 100 % opening of the motorized pump suction valve is confirmed, the motorized discharge valve begins to open.
Once 5 % open to be confirmed, the motorized valve of the pump discharge starts the motor.
Then the discharge and pipe pressure begin to be controlled through the 24 'control valve.'
Once the engine starts, it will take approximately 10 minutes for the pre-lubricator pump to turn off, which will activate again if the engine stops or there is low lubrication pressure in the engine.
After all the previous steps, the PLC begins to monitor the equipment variables coming from the field.
During pump operation, the system will continue to monitor the following variables.
Electrical characteristics of the motor OK. (Electric motor available)
Fully open motorized suction valve 01UV-0041A
The fully open motorized discharge valve 01UV-0041C
The differential pressure on the pump suction strainer, 01PDT-0041A, does not exceed 5 psi
Suction pressure 01PT0041A
Discharge pressure 01PT-0041B
Very high pressure at the discharge of the pump 01PSHH-0041 not active
Very low pressure at suction of the pump 01PSLL-0041ªv
Pump flow rate 01FQIT-0041
Forced-operation lubrication system
Operational Vibration Monitoring System
Operating temperature protection system
Monitoring of electrical parameters, vibration, temperatures, and processes on the motor/pump assembly must be maintained throughout the time the pump remains in operation.
3.3.2 Integration and design of the new technological architecture
Once the control system that satisfies the necessary requirements has been derived to guarantee the monitoring and control of the processes that are carried out in the Naughty Tank Farm, as well as the type of communications network and the necessary flex I/O modules, it was possible to proceed with the design of the technological architecture. It should be considered that control system 1 was selected, which is a modular system that operates under the AB interchange protocol through the 1756-ENET/B Ethernet card.
3.3.3 PLC control logic design that complies with the new specifications presented.
Main Routine: MainProgram. This is the main routine of programming; this means that all the other routines that make up the control program are called from the main one so that they can be executed. In addition, it establishes the order in which each routine must be executed to maintain proper control of the process in PTT. (See Figure 2).
Subroutine: DELETE_REGISTERS. It starts from rung 0 to rung 40, and is aimed at erasing all the PLC fault records, faults that originate specifically in the pumps, to then execute the commands that enable or disable the sequence of one or more pumps. Among the possible failures presented in the records are the following (taking pump 1 as a reference): Shutdown due to pump 1, shutdown due to lubrication pump off pump 1, shutdown due to fan off pump 1, shutdown due to suction closing traffic pump 1 discharge, stop due to electrical failure pump 1, stop due to low suction pressure pump 1, stop due to high discharge pressure pump 1, stop due to pump 1 on/off failure, stop due to pump 1 vibration. (See Figure 3).
Subroutine: GENERAL TANKS. It is in charge of enabling and disabling the receipt and dispatch sequences of the tanks, as well as deleting the records of the selected variables for the receiving and dispatch tanks. In the same way, there is the programming to make the stops in the receipt and dispatch of the tanks either by high-level or low-level switches. Among the commands used in this subroutine are the following (taking tank-9751 as a reference): low-level tank-9751, low-level tank-9751, stop by switch dispatching tank-9751, enable/disable dispatch sequence, high-level tank-9751, tank-9751 high level, stop by tank-9751 receipt switch, enable/disable receipt sequence, among others. (See Figure 4).
Subroutine: DISPATCH_TANKS. Dispatch activates the command to open the valves corresponding to the tank that is aligned to execute this operation. Some commands that are executed in this subroutine taking a tank (TK-9751) as reference are the following: val.mult.up transit open TK-9751, val.mult.up open TK-9751, high dispatch TK9751, val.mult.under transit open TK-9751, val.mutl.under open TK-9751, office under 9751, among others. See Figure 5.
Subroutine: PUMP_1. The permissions to start the pump remotely are verified, as well as the availability to start the pump, and gradually after executing the start command, the program verifies the lubrication level of the pump, the fan status, the status of the suction and discharge valves, and in the event of a failure in one of these components, the order to stop pumping by this pump is automatically executed. Among the commands used in this subroutine are the following: local remote fan B1, local remote lubricator pump 1, local remote pump 1, automatic manual pump 1, status of fan B1, ind.l/r act. pump 1 suction, pump 11 discharge ind.l/r, remote permitting, start availability. See Figure 6.
3.3.4 Design of the control and supervision schematic in SCADA guardian alba of the output manifold.
To perform the 3D modeling of the output manifold, the blender software was used. This design can be seen in the following image. See Figure 7.
In order not to compromise operational continuity, all control routines developed were thoroughly evaluated (logic review, interlocks, alarms, communications, and fault handling). Only after passing these checks were, they loaded into the PLC and run exclusively in an isolated test environment with simulated I/O and replicated plant conditions, in order to validate their operation and safety without affecting operation.
The advance of technology in recent times has given another perspective to industrial processes, displacing arduous human labor with many opportunities for the supervision of computerized processes (Viraktamath et al., 2020; Acharya et al., 2017; Chakraborty et al., 2020). This transformation is largely driven by the integration of industrial automation technologies, which program systems to perform tasks autonomously and monitor operational variables to ensure system stability and performance (Tomar & Kumar, 2020; Tomar et al., 2023). These technologies increasingly rely on artificial intelligence algorithms to make real-time decisions, reduce uncertainty, and dynamically adjust process parameters particularly in energy-intensive industries such as oil and gas.
Automation is a set of technologies that allow machines and systems to work without the continuous intervention of a human operator, achieving superior performance compared to manual operations (Inzunza et al., 2020; Mendoza, 2016; Abdu & Taleb, 2014; Ibrahim et al., 2018). In particular, Tomar et al. (2023) emphasized the growing role of AI-based industrial control in optimizing output while reducing human error and downtime. Similarly, Kiangala & Wang (2019) proposed the concept of intelligent manufacturing systems, highlighting the shift toward cyber-physical production environments that leverage automation, data exchange, and real-time feedback.
The industrial processes of today are oriented toward achieving continuous improvement in operational reliability, identifying opportunities for enhancement through automated maintenance systems and predictive diagnostics (Nadgauda & Muthukumaraswamy, 2019; Hussein et al., 2022). In this context, Li et al. (2021) demonstrated that predictive and adaptive control strategies in pipeline systems can significantly reduce energy consumption and flow irregularities, reinforcing the importance of advanced control architectures in complex industrial networks.
Currently, the trend in the field of automation and process control aims to improve operations carried out by various manufacturing companies, including batch and continuous production, by replacing existing manual systems with fully self-controlled and reconfigurable systems (Alphonsus & Abdullah, 2016; Yu & Xue, 2022). Specifically, oil companies characterized by continuous production require uninterrupted monitoring and control across most of their operations (Díaz, 2019; Pino, 2013). This allows them to optimize throughput while avoiding incidents that may cause operational disruptions or safety risks (Ronceros et al., 2023; Tomar et al., 2022). A notable example is the automated control system for three-phase separators developed by Wu et al. (2022), which improved dehydration efficiency by dynamically adjusting flow and pressure conditions, reducing water content in crude oil to optimal levels.
Below is a summary table of the expected benefits from a technical, operational, and economic perspective of implementing the proposal. The percentages are indicative estimates based on functional modeling of the case and benchmarks from the literature and comparable automation practices. Table 1 presents the expected benefits across perspectives along with estimated improvements to be validated during implementation.
4. CONCLUSIONS
An analysis of PTT's pump room operations and associated process equipment led to a proposal for an automated outlet manifold control system at PTT's Crude Oil Storage and Transportation Center. The system is designed to optimize crude oil pumping and maintain sufficient pressure to ensure supply from the tank farm to the José storage and shipping terminal, the Guaraguao storage and shipping terminal, and the Anaco crude oil storage and transportation center, facilitating the achievement of planned production targets in terms of volumes, deadlines, and quality levels required by customers.
The proposed system enables 100 % automated control of the discharge pressure in the main PTT pumps and pipelines, a process that regulates the flow of crude oil transported, allowing the operator to monitor the information transmitted in the control room and make the necessary adjustments to maintain the corresponding parameters in the valve management, and thus control the flow and pressure to achieve an optimal delivery of crude oil.
The proposed control logic would reduce the need to dispatch operators to high-risk areas by enabling automatic start/stop of pumps in the PTT pump rooms. Based on functional modeling and cycle-time estimates, the expected improvement in response time is approximately 60-80%. By contrast, manual discharge-pressure control using two-position isolation valves (open/closed), sometimes operated in a quasi-throttling manner, subjects the system to elevated pressure transients and mechanical loads, increasing risks to personnel safety and equipment integrity
The study can be replicated in other multiple outlets of crude oil tank yards of oil industries that use crude oil pumping processes considering the nature and particularity of its platform, which could generate the use of the proposed control logic with some adjustments oriented to the equipment and platform of the oil company.



























