Refining India September 2025 Issue

FV07

STRM10

STRM12

STRM07

FV11

STRM11

PIC03

HEX01

CND01

PIC02

STRM08

LIC03

PSP01

STRM06

FV06

Dibenzoyl peroxide

STRM02

VLV02

FV01

STRM09

FV04

STRM04

STRM01

PIC01

Methyl methacrylate Butyl acrylate Styrene Acrylic acid 2-mercaptoethanol Solvent

Solvent

LIC02

REC01

TIC01 LIC01

FIC03

MTK01

VLV01

FV05

STRM05

STRM03

Thick resin

Finished product

Figure 4 Adding control system to the model

be added to the model similarly, by laying this new information over the existing model. AI companion can then access the information in this model and reason using it to perform many types of analyses. Capabilities of the AI companion The current AI companion has a chat interface where the user can ask a question, and the AI technology then answers the question. It accesses the enterprise AI model of the plant to get the relevant information and uses this information to get to the answer. The AI companion demonstrated the following capabilities using the enterprise AI model of the plant:  It can identify the equipment in a plant and can understand the processes taking place inside each equipment.  It can understand the flow of each stream and its relationship with the equipment.  It can analyse the impact of flow deviations on upstream and downstream equipment.  It can analyse and identify the cause of any process upset and its consequences. HAZOP study by AI companion The AI companion can utilise the capabilities described in the previous section to carry out a HAZOP study of the plant. It can identify the

parts relevant to HAZOP study and then apply a guide word to the design intent. It can then deduce the causes and consequences of such a deviation, as shown below:  Identification of the relevant parts for HAZOP study: Can accurately identify the equipment, streams, sensors, and valves as the relevant parts (nodes) of the plant that should be studied for the HAZOP.  Applying a guide word to the design intent of a plant: Can apply a guide word to the design intent of the plant. It also deduced the causes and consequences of such a deviation. In the example shown in Figure 5 , the design intent was defined as ‘the flow rate of process stream STRM04 should not differ from the flow rate of process stream STRM03 by more than 5%’. AI companion was then asked to apply the guide word ‘less’ to the design intent and identify the causes and consequences of this deviation. It identified all relevant causes for the flow rate of the process stream STRM04 to be less than the design intent. Similarly, it deduced the consequences of this deviation.  Identifying existing safeguards in the process to prevent the deviation : Can identify the control mechanisms existing in the plant, which help prevent the deviation.  Suggesting additional safeguards: Can identify equipment and systems missing from the plant,

Refining India

56

Powered by