IEC TR 62443-2-3:2015, "Security for industrial automation and control systems — Part 2-3: Patch management in the IACS environment," is the part of the IEC 62443 series that addresses one of the hardest operational problems in OT security: how an asset owner and an IACS product supplier actually exchange, test, approve and track patches without breaking a running plant. Published in 2015 as a Technical Report rather than a normative International Standard, it describes a patch lifecycle model, recommended requirements for both sides of the relationship, and a structured data-exchange format — filling a gap that sits between IEC 62443-2-1's asset-owner security program and the vulnerability-handling expectations referenced across the rest of the series.
Key takeaways
- A failed patch on a control system can stop a process or hurt someone, not just break a laptop.
- Clause 4.4 and Table 1 of the IEC preview text define eleven states: four belong to the product supplier, five to the asset owner, and two (Available and Released) can be managed by either side.
- Clause 5 asks the plant to know, for every device that can be updated, what is installed, what exists and what the supplier has cleared.
- A device maker should ship enough information for the asset owner to move each patch from Released to Authorized without guessing.
Why is patching an IACS a different problem than patching IT?
A failed patch on a control system can stop a process or hurt someone, not just break a laptop. The Introduction frames the issue directly: many organizations already run well-established information security management systems under ISO/IEC 27001 and ISO/IEC 27002, but applying that IT-style cyber security strategy to an IACS "without understanding the consequences" can cause harm — "successful attacks against industrial systems may lead to health, safety and environmental (HSE) consequences" that a pure IT patching program never has to weigh.
Clause 4.1 catalogs the practical obstacles: preparing an IACS to be patched "can require a tremendous amount of work," since each patch for each product means gathering and analyzing patch information, installing and verifying on a test system, backing up before and after, retesting, and documenting the change. Because of that overhead, most organizations schedule patching around existing maintenance outages — sometimes quarterly, yearly, or less often — and "some extremely critical systems may not have outage windows available and can therefore not be patched if a system outage is required to do so." The report is explicit that applying a patch is a risk-management decision: "if the cost of applying patches is greater than the risk evaluated cost, then the patch may be delayed," particularly where other compensating controls already mitigate the risk.
Clause 4.2 states the consequence of getting this wrong plainly: poor IACS patch management increases "the risk of loss or compromise of an IACS system," and unlike office systems, that compromise "may impact system safety, the physical safety of operational personnel, the quality of produced products, the safety of produced products and the usability of produced products." Clause 4.3 adds a second, longer-running problem: IACS equipment often stays in production "for decades," so asset owners regularly face patches that no longer exist for obsolete, unsupported hardware and need other mitigations when patching simply isn't an option.
What are the eleven patch states, and who owns each one?
Clause 4.4 and Table 1 of the IEC preview text define eleven states: four belong to the product supplier, five to the asset owner, and two (Available and Released) can be managed by either side.
| Patch state | Definition (per Table 1) | Managed by |
|---|---|---|
| Available | Provided by a third party or IACS supplier but still untested | Asset owner / Product supplier |
| In Test | Being tested by the IACS supplier | Product supplier |
| Not Approved | Failed supplier testing; not to be used unless later Approved | Product supplier |
| Not Applicable | Tested and found not relevant to IACS use | Product supplier |
| Approved | Passed supplier testing | Product supplier |
| Released | Released for use by the supplier or third party, or directly applicable to internally developed systems | Asset owner / Product supplier |
| In Internal Test | Being tested by the asset owner's own testing team | Asset owner |
| Not Authorized | Failed internal testing, or found not applicable | Asset owner |
| Authorized | Released by the asset owner and meets internal standards (or didn't need testing) | Asset owner |
| Effective | Posted by the asset owner for use | Asset owner |
| Installed | Installed on the system | Asset owner |
Figure 1 draws this as a state-transition diagram: the supplier's states sit in the dark-gray left half, feeding into the asset owner's light-gray right half (In Internal Test → Authorized/Not Authorized → Effective → Installed). Knowing which state every patch for every device currently sits in — not just whether it has been "installed" — is the document's core operational idea.
What does Clause 5 ask asset owners to do?
Clause 5 asks the plant to know, for every device that can be updated, what is installed, what exists and what the supplier has cleared. It frames this as an "implied obligation to uphold the safety, reliability, operability, security and quality" of operations. IACS asset owners should:
Maintain an inventoryof every device that can be updated — modified in functionality, configuration, operation, software, firmware or operating code — referred to as "updatable" devices. Record the installed versioncurrently running on each device. Determine, on a regular schedule, the latest versionavailable for each device. Determine the released versionsthe IACS product supplier has identified as compatible and that meet the asset owner's own standards. Test patch installationin a way that "accurately reflects the production environment," so reliability and operability aren't harmed once the patch reaches the real system — patches that pass become "authorized." Schedule authorized, effective patchesfor installation "at the next available opportunity," respecting system-design constraints (redundancy, fault tolerance, safety) and operational ones (scheduled or unscheduled outage windows, on-process operation). Update records at least quarterly, covering installed, authorized, effective and released versions for every updatable device.
Clause 6 mirrors this with a shorter list for the product supplier, and Clause 7 defines how the two sides exchange patch information. Annex B (asset-owner guidance, pages 26–55 of the report) and Annex C (supplier and service-provider guidance) carry the step-by-step detail; their wording is in the purchased text on the IEC webstore.
What should a device maker ship with every patch?
A device maker should ship enough information for the asset owner to move each patch from Released to Authorized without guessing. Items 2, 4 and 5 of Clause 5 only work if the supplier publishes version, compatibility and test information. On the supplier side, the IEC 62443-4-1:2018 contents list Practice 7, "Security update management", with five requirements: SUM-1 Security update qualification, SUM-2 Security update documentation, SUM-3 Dependent component or operating system security update documentation, SUM-4 Security update delivery and SUM-5 Timely delivery of security patches. The table pairs each asset-owner need with a deliverable; the deliverable column is SoftWSP's reading, not text from either standard.
| Asset owner needs (2-3 Clause 5 / Annex B) | What the device maker ships | Closest 4-1 practice |
|---|---|---|
| Released, compatible versions (item 4) | Release notes naming the affected product, hardware revisions and the firmware versions the patch replaces | SUM-1, SUM-2 |
| A test that mirrors production (item 5, B.6) | Supplier test summary: what was regression-tested, on which configuration, with known side effects | SUM-1 |
| Patch file authenticity (B.6.3) | Signed image plus published hash, and how the device verifies the signature at install | SUM-4 |
| Roll back and restore (B.6.7) | Whether downgrade is allowed, how to restore configuration, minimum safe version | SUM-2 |
| Third-party components (B.5.3) | Which open-source or OS updates the patch includes, so CVEs in those parts can be closed | SUM-3 |
| An outage plan (item 6, B.8.4) | Restart scope and expected downtime per device | SUM-2 |
| Risk mitigation when no patch exists (4.3, B.6.8) | Workaround or mitigation text and a date for the fix | SUM-5 |
A component SBOM for each release covers the third-party row cheaply; see SBOMs for industrial automation under IEC 62443 and the IEC 62443-4-1 secure development lifecycle guide.
What is the VPC exchange format, and what do suppliers publish instead?
VPC ("Vendor Patch Compatibility", Clause 7 and Annex A) is a 2015 XML/XSD file that tells the asset owner which patches are compatible with which supplier products; most ICS suppliers now publish the same facts in CSAF advisories. Clause 7 adds a filename convention, and Table 2 through Table 5 define the PatchData, PatchVendor, Patch and VendorProduct elements. IEC has not revised 2-3 since 2015: the webstore record shows edition 1.0, a stability date of 2027 and no project under development. CISA, for example, publishes its ICS advisories as machine-readable CSAF 2.0 files on GitHub.
CSAF 2.0 covers most of what VPC carried, plus fields 2-3 asks for in prose:
| What 2-3 wants exchanged | CSAF 2.0 field |
|---|---|
| Which product versions a vulnerability affects | product_status → known_affected, first_affected, last_affected |
| Which version carries the fix | product_status → fixed, first_fixed, recommended |
| Vendor fix, workaround or mitigation | remediations[].category: vendor_fix, workaround, mitigation, or none_available when no fix is available |
| Outage planning | remediations[].restart_required.category, from none through service, system, machine and zone |
| Not relevant to this product | product_status → known_not_affected |
What CSAF does not carry is the asset owner's own state (In Internal Test, Authorized, Effective, Installed); that stays in the plant's records. See CSAF/VEX for device makers for authoring.
What does one firmware patch look like moving through the lifecycle?
The example below is synthetic: the company, product and versions are invented to show the state changes, not taken from a real advisory.
AvailableA TLS library CVE is published. "ExampleCo" learns its X200 controller firmware 4.2.1 ships the affected library version. In Test → ApprovedExampleCo builds 4.2.3 with the patched library, runs its regression suite on the X200 hardware revisions B and C, and finds revision A cannot take the update (Not Applicable for that revision). ReleasedExampleCo publishes a CSAF advisory: 4.2.1 known_affected, 4.2.3fixed, avendor_fixremediation withrestart_required=machine, and a firewallworkaroundfor revision A. It ships the signed image, hash and release notes.In Internal Test → AuthorizedThe plant installs 4.2.3 on a spare X200 in its test cell with the production program loaded, checks scan times and I/O, and authorizes it. EffectiveThe patch is posted to the plant's approved software store. InstalledTwo of six controllers are updated in the next outage; four stay on 4.2.1 behind the workaround until the next window. The quarterly record (Clause 5 item 7) shows 2 Installed, 4 Effective-but-not-installed.
Where does 2-3 sit in the IEC 62443 series?
62443-2-3 is a Technical Report in the "Policies and procedures" (2-x) group, alongside 2-1 (asset owner security program, current edition 2024), 2-2 (published as IEC PAS 62443-2-2:2025 on 2025-03-11) and 2-4 (requirements for IACS service providers). Its two normative references are IEC TS 62443-1-1 (terminology, concepts and models) and 2-1, which 2-3's patch activities feed into as a supporting process. 2-3 deliberately stays out of vulnerability-disclosure ethics and process — that ground is covered by ISO/IEC 29147 and 30111 (see the CVD process explained) and by each asset owner's own IACS security management system. For the wider OT program, see the NIST SP 800-82 OT security guide.
Which mistakes come up most often?
Most failures come from treating the lifecycle as two states, patched and unpatched.
- No updatable-device inventory. Clause 5's tracking is impossible without a baseline list of every device that can receive a patch; the same list feeds IEC 62443-3-2 risk assessment and zone partitioning.
- Authorized counted as installed. A patch that passed internal test is not deployed; conflating the two hides how much of the fleet is current.
- No plan for systems without an outage window. Clause 4.1 says some critical systems have none, so the mitigation question (Clause 4.3, Annexes B.4.5, B.4.6, B.8.5) has to be answered before a vulnerability is disclosed.
- Release notes without versions. "Security improvements" with no affected-version range forces the plant to test blind.
- Silent third-party updates. A firmware patch that bumps an OS or library without listing it leaves CVEs open on paper.
- Expecting 2-3 to cover disclosure. Notes 1 and 2 to its scope exclude vulnerability disclosure and the gap before a patch exists; pair it with your CVD process.
Which parts can a firmware or network scan cover?
A firmware scan covers the supplier's component list and a network capture shows which hosts talk; neither tracks the plant's patch states. On the supplier side, SoftWSP's firmware analysis extracts a CycloneDX 1.6 SBOM from a firmware image and matches its components against NVD, CISA KEV, FIRST EPSS and CISA CSAF ICS advisories; AI drafts a cited report that an engineer reviews. It does not write VEX or CSAF statements and does not analyse reachability. On the plant side, OT Snapshot works from Zeek and ICSNPP logs of a network capture and lists which hosts speak Modbus, S7comm, DNP3, EtherNet/IP or BACnet. It does not read installed firmware versions, so Clause 5's "installed version" column still comes from the asset owner's records. Once an engineer has decided a product's status, the free VEX generator writes it as CSAF 2.0 in the browser; for firmware work see device firmware security.
This guide quotes IEC's published preview of IEC TR 62443-2-3:2015 (foreword, introduction, scope, definitions, Clauses 4 and 5, contents) and the IEC webstore record. Clause 6, Clause 7 detail and Annexes B and C are in the purchased standard. It is not compliance or legal advice.
Put this guide to work
Frequently asked questions
What is IEC 62443-2-3?
IEC TR 62443-2-3:2015, "Security for industrial automation and control systems — Part 2-3: Patch management in the IACS environment," is a Technical Report (not a normative International Standard) in the IEC 62443 series. Per its own Clause 1 scope text, it "describes requirements for asset owners and industrial automation and control system (IACS) product suppliers that have established and are now maintaining an IACS patch management program," and "recommends a defined format for the distribution of information about security patches from asset owners to IACS product suppliers, a definition of some of the activities associated with the development of the patch information by IACS product suppliers and deployment and installation of the patches by asset owners." It was published 30 June 2015 as a first-edition Technical Report, 61 pages, prepared by ISA Technical Committee 99 in partnership with IEC Technical Committee 65, carrying the ISBN 978-2-8322-2768-8. The IEC webstore lists its stability date as 2027 and its status as current, not withdrawn.
Why is it a Technical Report, and does it cover non-security patches too?
The foreword explains that while IEC technical committees mainly prepare International Standards, a committee may instead publish a Technical Report when it has collected "data of a different kind from that which is normally published as an International Standard, for example 'state of the art'" — 2-3 documents patch-management practice and a data-exchange format rather than setting testable conformance requirements. Its scope clause states the exchange format and activities "are defined for use in security related patches; however, it may also be applicable for non-security related patches or updates," and explicitly "does not differentiate between patches made available for the operating systems (OSs), applications or devices" or between suppliers of infrastructure components versus IACS applications — it covers all patch types (bug fixes, reliability or operability fixes, and security fixes) uniformly. Two things it explicitly excludes, per its own Notes 1 and 2: guidance on the ethics or process of vulnerability discovery and disclosure, and guidance on mitigating a vulnerability's risk in the window between disclosure and patch availability (that is left to other parts of the 62443 series, including the asset owner's own IACS security management system under [IEC 62443-2-1](/guides/iec-62443-2-1-asset-owner-security-program/)).
What is the patch lifecycle model, and who manages each state?
Clause 4.4 and Table 1 define an 11-state lifecycle, split between states the IACS product supplier manages and states the asset owner manages. Supplier-managed states: In Test, Not Approved ("failed the testing of the IACS supplier"), Not Applicable ("tested and is not considered relevant to IACS use"), and Approved ("passed testing by the IACS supplier"). Two states are shared (asset owner or supplier): Available ("provided by a third party or an IACS supplier but has not been tested") and Released. Asset-owner-managed states: In Internal Test, Not Authorized ("failed internal testing, or may not be applicable"), Authorized ("released by the asset owner and meets company standards for updatable devices, or by inspection did not need testing"), Effective ("posted by the asset owner for use"), and Installed ("installed on the system"). Figure 1 shows the full state transition diagram; the document notes "not all patches available are relevant to the IACS and not all patches are compatible with the IACS applications," which is why tracking every patch's current state matters.
What does IEC 62443-2-3 recommend asset owners actually do?
Clause 5 lists recommended requirements, framed as an implied obligation to uphold "the safety, reliability, operability, security and quality" of operations. Verbatim, IACS asset owners should: (a) maintain an inventory of all 'updatable' devices; (b) record the currently 'installed' version for each device; (c) determine, on a regular schedule, the 'latest' available version for each device; (d) determine the 'released versions' the IACS product supplier has identified as compatible; (e) test patch installation in a way that 'accurately reflects the production environment' before calling a patch 'authorized'; (f) schedule authorized, effective patches for installation 'at the next available opportunity' within system-design and operational constraints; and (g) update these records at a defined interval, at least quarterly. Clause 6 sets a parallel, shorter set of recommended requirements for the IACS product supplier side (this guide's source preview text runs through clause 5g; see the IEC webstore link below for the purchased text covering Clauses 6-7 and the Annex B/C implementation guidance in full).
What is the VPC exchange format, and is it still relevant?
Clause 7 and Annex A define "Vendor Patch Compatibility" (VPC) — an XML/XSD-based file format for a supplier to tell an asset owner, in a structured way, which patches are compatible with which of the supplier's products. Table 2 to Table 5 specify the PatchData, PatchVendor, Patch and VendorProduct file elements. This is the document's most dated section: it predates CSAF/VEX-style machine-readable advisory formats that have since become the more common way vendors publish patch-applicability data (see [CSAF/VEX for device makers](/guides/csaf-vex-for-device-makers/)). IEC has not withdrawn or revised 2-3 to reconcile the two; a 2015-era OT shop evaluating exchange formats today should treat the VPC schema as one historical option alongside CSAF/VEX rather than the default choice.
Sources
- IEC: IEC TR 62443-2-3:2015 — Security for industrial automation and control systems — Part 2-3: Patch management in the IACS environment (webstore record: edition 1.0, 61 pages, TC 65, stability date 2027)
- IEC TR 62443-2-3:2015 — Document Preview (foreword, introduction, full scope, definitions, Clause 4 patching problems and patch lifecycle Table 1/Figure 1, Clause 5 asset-owner requirements; iTeh Standards sample PDF, reproduced here under IEC's short-quotation allowance)
- ISA: The ISA/IEC 62443 Series of Standards (series overview, including Part 2-3 "Patch management in the IACS environment")
- Wikipedia: IEC 62443 (series structure, part-by-part summary; cross-checked against the IEC and ISA pages above for this guide)
- IEC 62443-4-1:2018 — Document Preview (contents: Practice 7 Security update management, SUM-1 to SUM-5; iTeh Standards sample PDF)
- IEC: IEC PAS 62443-2-2:2025 — webstore record (Publicly Available Specification, published 2025-03-11)
- OASIS: Common Security Advisory Framework (CSAF) Version 2.0, OASIS Standard (product status, remediation categories, restart_required)
- CISA: CSAF repository (machine-readable ICS advisories in CSAF 2.0)
About this guide
SoftWSP OT/ICS Desk
Covers the OT/ICS security cluster: 75 guides.

In this cluster
Questions about this page?
