Reading Legacy 1980s P&IDs. A Digitizing Guide.

The brownfield retrofit projects an industry runs on ride on legacy P&IDs. A typical refinery's drawing vault has hundreds of original 1980s drawings, decades of marked-up redlines, occasional CAD-conversion attempts that never reached as-built status, and the controlled drawing, which is whatever the MOC process declares is the current revision.

Reading those drawings is half the work of a brownfield project. Doing it cleanly avoids weeks of forensic investigation later.

What the older drawings get right

Pre-1990s P&IDs read better than most modern ones. They were drafted by hand on vellum or mylar by drafters who specialized in process drawings, and the line work, lettering, and layout were treated as craft. The clarity is something CAD-generated drawings struggle to match because the human drafter could vary line weight, tag placement, and crowding to communicate intent.

What you can trust on a 1980s drawing.

• Tag identification follows ISA 5.1. First-letter, second-letter conventions are old. F means flow on a 1985 drawing. T means temperature.
• Equipment numbering is plant-specific but stable within the plant. The drafter's loop-numbering convention has been preserved through three decades of revisions on most plants.
• Process intent is documented in the notes block. Drafters of the era wrote more in the notes section because they could. CAD-generated drawings ship with terse or auto-generated notes.

Legacy notation reference

The gap between ISA-S5.1-1984 and modern ISA 5.1-2022 is mostly symbolic. The first-letter, modifier, second-letter, subsequent-letter tag structure is the same. The following areas are where older drawings diverge enough to cause confusion.

Instrument symbol shapes. Modern ISA 5.1 uses a circle for every field instrument, a circle with a horizontal line through it for instruments mounted in a panel, accessible to the operator, and a circle with a dashed horizontal line for instruments behind the panel, not directly accessible. Older drawings, especially from North American engineering firms in the early 1980s, sometimes used a square with a line through it for panel-mounted discrete instruments and a rectangle with internal annotation for rack-mounted multi-loop controllers. If you see rectangular instrument symbols, they are almost certainly representing a panel-mounted discrete controller or indicator in the old convention.

Instrument rack notation. Multi-loop controllers in the 1980s were physical rack hardware, Foxboro Spec 200, Taylor Mod 30, Honeywell TDC 2000 modules. Drawings would show a shared rack symbol, typically a rectangle spanning several loops, with individual function blocks inside it. The rack rectangle is not a vessel or a PLC. It is the control hardware. Engineers unfamiliar with this convention sometimes misread rack outlines as process equipment.

Pneumatic loop hash marks. Older drawings that predate the all-electronic era used a diagonal hash mark crossing the signal line between a pneumatic transmitter and its receiver to denote a pneumatic signal, 3-15 psi. The hash mark distinguishes the pneumatic line from an electrical wire. If you see hash marks on signal lines and no corresponding ISA 5.4 signal-type annotation, assume pneumatic. On plants that converted from pneumatic to electronic in the 1990s, some drawings show both the original hash-marked pneumatic line and a new solid line for the 4-20 mA replacement, with a handwritten note indicating the cutover date.

Older valve symbols. Gate valves on 1980s drawings are sometimes shown as two triangles point-to-point, identical to a globe valve representation in some later standards. Control valve actuator symbols varied by engineering firm. Some firms used a circle above the valve body for a diaphragm actuator. Others used a rectangle. The drawing's legend sheet is always the authority for the specific firm's convention. If the legend sheet is missing, common on fragmented brownfield drawing sets, identify the engineering firm from the title block and look up their standard convention for the era.

Crossed-line vs jumped-line conventions. On a drawing with many crossing signal lines, the convention for "these lines cross but do not connect" was either a small arc, jump over one line, or a small gap in one line. The modern ISA convention uses the jump arc. Older drawings from some firms used neither. Two crossing lines were assumed to not connect unless a dot was drawn at the intersection. If you are reading a 1980s drawing from a firm that used the no-symbol convention, every line intersection is potentially ambiguous. Context, does connecting these two signal lines make process sense. Resolves most cases.

Mapping ISA-S5.1-1984 to ISA 5.1-2022

The 1984 edition and the 2022 edition share the same tag-identification logic. The differences are additive. The 2022 edition adds symbols for digital communications, fieldbus devices, wireless transmitters, and Safety Instrumented System functions that simply did not exist in 1984.

For instrument type codes, first and subsequent letters, the mapping is effectively one-to-one for all conventional measurement and control tags. FIC, PIT, TT, LCV, and similar codes mean the same thing in both editions.

For instrument location, the horizontal line convention, treat older drawings by looking at the legend first, then applying the closest modern equivalent.

The hexagon for DCS functions is the single largest source of confusion on legacy drawings. A hexagon in the 1980s represents any function running inside the distributed control system. By 1990, expanded conventions distinguished between shared-display, shared-control functions, the hexagon and computed functions, programmable logic controllers, and data links. On a 1985 drawing, a single hexagon may represent what a modern drawing would show as four separate function blocks.

Drafting-shop conventions of the era

Major control-system vendors of the era had published application guides that their customers' engineering firms treated as drawing standards. Two are common enough to recognize on sight.

Foxboro Spec 200. Drawings produced for plants with Spec 200 control systems, 1970s through mid-1980s often use Foxboro's proprietary module designations inside the instrument bubbles. The function codes correspond to Foxboro module types, such as a ratio controller or a derivative unit, not to ISA first-letter, subsequent-letter codes. If you see unfamiliar letter combinations inside bubbles on a drawing with a Foxboro title block or a Foxboro module list on the legend sheet, consult the Spec 200 module catalog to map module codes to ISA equivalents.

Bailey Net 90. Bailey Controls, later ABB Net 90 systems were common in power generation and some chemical plants from the late 1970s through the 1980s. Bailey drawings use their own function-block notation for control logic, and the P&ID may show Bailey module names, AFC, APC, DFC inside or adjacent to ISA circles. These are the DCS function blocks. Map them to the ISA Y, computing relay convention when digitizing.

Missing tag prefixes. Some operating companies in the 1980s omitted tag prefixes that are standard today. K-prefix sequencer tags were widely dropped because the sequencer logic moved into PLC code. HH, LL alarm tags were sometimes not drawn on the P&ID, with only a note referencing a set-point table. If you are extracting an instrument index and the drawing clearly shows an instrument with only partial tag information, a loop number but no instrument type letter, or a letter with no number, the missing data is either in the notes block, in a separate set-point or alarm schedule, or was lost in a redrafting pass. Do not fabricate the missing field. Flag it for engineering resolution.

How to digitize them cleanly

Scan at 300 plus dpi. Lower resolution misses fine line work and small annotations. High-resolution scanning is inexpensive. The cost of missing tags is high.

Capture the redlines as part of the as-built. If the plant operates on the redlined drawing, the redlined drawing is the as-built. Re-drafting from the original-as-drafted strips field changes that have been operational for decades.

Run extraction against the scanned drawings, not against a re-typed Excel inventory. Extraction tools read what is on the drawing, tag, location, signal class and produce a structured dataset. A re-typed Excel inventory captured by a previous engineer is only as good as the engineer's transcription discipline. The brownfield engineering guide covers the full workflow from scanning through reconciliation and what to do with the residual discrepancies.

Cross-reference against the legacy I/O list as a separate step. The discrepancy report between drawing-extracted scope and Excel-list scope tells you where field changes have outpaced the documentation, where the I/O list owner has already caught up, and where the drawing is missing instruments that are operational in the field.

Walk the plant for the residue. No drawing extraction produces a perfect dataset on its own. The last 5 percent of accuracy on a 30-year-old plant comes from a senior engineer walking the plant against the extracted dataset and confirming each anomaly.

Common digitization errors on legacy sets

These days the practical approach is to scan first, extract second, reconcile against whatever I/O list the operations team is keeping current, and walk the plant last for the residual handful that can't be settled from documents. Same end result. Days instead of weeks.

The drawings from the 1980s are not the problem. The problem is treating them as if they were the design intent rather than the operational reality. Read what's there.

Related

• Brownfield P&ID digitization
• Scanned P&IDs vs CAD exports
• ISA 5.1 letter codes
• P&ID revision comparison
• Commissioning loop check plan

FAQ

Are 1980s P&IDs still valid as controlled documents.

If they are the as-built drawing for an operating plant and they have stayed inside the operating company's MOC process, yes. Many operating units run on as-built drawings that are 30-40 years old, with field changes layered on as marked-up redlines. The drawings are valid. They may just be hard to read for engineers raised on AutoCAD.

What ISA 5.1 conventions changed between the 1980s and now.

Tag identification rules are largely the same. Symbol conventions for shared-display and digital devices are newer. Pre-1990s drawings use simpler conventions because DCS systems were either not present or represented with a single hexagon. Computing relay, Y annotations are clearer in modern editions. Most differences are stylistic rather than semantic.

Can extraction tools read scanned 1980s drawings.

Yes, if the scan quality is sharp and the drawing was originally drafted to a recognizable convention. Tools handle ink-on-vellum drawings, microfilm reproductions, and modern scans equally well as long as the tags are legible. Hand-marked redlines extract too. The tool reads what is on the drawing today, not what was originally drafted.

How do I handle a drawing where some instruments have Foxboro or Bailey function codes instead of ISA tags.

The vendor function codes inside bubbles on Foxboro Spec 200 or Bailey Net 90 drawings are DCS module designations, not ISA subsequent-letter codes. When extracting, log the bubble contents as found, then add a reconciliation column mapping each vendor code to its ISA equivalent using the vendor's module catalog. Do not attempt to auto-map these codes without the catalog. A Foxboro AFC and an ISA FFC are not the same thing.

My drawing set has no legend sheet. How do I know which conventions apply.

Check the title block for the engineering firm and the approximate drawing date. Most North American firms in the 1980s followed ISA-S5.1-1984 for tag identification and their own house standard for symbol shapes. If you cannot identify the firm, treat circles as the universal instrument symbol and resolve any ambiguous rectangular symbols by checking whether they connect to process lines, likely instrument or span multiple loops without process connections, likely control rack. Flag all ambiguous symbols for engineer review rather than guessing.