Key takeaways
- ICCP's only native data-security mechanism is a bilateral table — a configuration object on each ICCP server that defines which specific data points a given remote peer is authorized to read or write.
- Public vulnerability records for ICCP are implementation bugs, not protocol flaws.
- No current NERC CIP standard names ICCP explicitly.
ICCP (the Inter-Control Center Communications Protocol, standardized as TASE.2 / IEC 60870-6) is how power-grid control centers, balancing authorities and grid operators exchange real-time telemetry, control commands and scheduling data with each other — and its base protocol was built in the early 1990s with no native encryption or authentication, relying instead on access-control tables configured at each ICCP server. A security extension, IEC 62351-4's "Secure ICCP" profile, has existed for over a decade but national-lab assessments and NERC's own 2026 reporting describe it as still rarely deployed across the installed base.
What is ICCP/TASE.2, and where does it run?
ICCP grew out of the Electric Power Research Institute's Utility Communications Architecture (UCA) program in the early 1990s, when a Utility Communications Specification working group submitted the protocol to IEC Technical Committee 57, which standardized it as the IEC 60870-6 series. The core document, IEC 60870-6-503 ("TASE.2 Services and protocol"), is described by IEC as specifying "a method of exchanging time-critical control centre data through wide-area and local-area networks," with companion documents covering the functional profile (60870-6-702), the object models that define what data ICCP can carry (60870-6-802), and transport profiles (60870-6-601 and TS 60870-6-602). Unlike field-level protocols such as Modbus or DNP3, which typically run between a control center and field devices, ICCP's job is control-center-to-control-center: exchanging real-time measurements, control operations, scheduling and accounting data, and event notifications between utilities, regional transmission organizations, independent system operators and non-utility generators that otherwise have no reason to trust each other's networks.
Why does the base protocol have no built-in security?
ICCP's only native data-security mechanism is a bilateral table — a configuration object on each ICCP server that defines which specific data points a given remote peer is authorized to read or write. EPRI's own 2001 assessment is direct about the limits of that mechanism: "ICCP provides access control via bilateral tables maintained on an ICCP server. This is the only data security service provided within ICCP." There is no protocol-level encryption and no cryptographic authentication of the peer at the other end of the link. Idaho National Laboratory's 2008 common-vulnerabilities report for control systems warned that "unencrypted ICCP communication... may provide many opportunities for an attacker to intercept, understand, and alter the data," and a decade later Pacific Northwest National Laboratory's PNNL-26729 (2017) assessment called ICCP "an unprotected protocol" exposed to integrity violation, interception or alteration, spoofing and eavesdropping by anyone able to reach the link.
What does IEC 62351 add, and is it actually used?
The fix exists on paper: IEC 62351-4 ("Profiles including MMS") defines security profiles specifically for ICCP and IEC 61850, layering the TLS-based transport security specified generically in IEC 62351-3 on top of the base protocol, along with certificate-based mutual authentication — an implementation the industry calls "Secure ICCP." Sandia National Laboratories documented early Secure ICCP integration work as far back as 2007 (SAND2007-3345). Yet PNNL's 2017 assessment still described Secure ICCP as "rarely implemented if at all," and NERC's own January 2026 Critical Infrastructure Protection Roadmap confirms the gap hasn't closed at the industry level, stating plainly that "legacy protocols, such as DNP3, ICCP, and Modbus often traverse unencrypted links that fall outside the current CIP-012 scope." In other words, the standards-based mitigation has existed for roughly two decades without becoming the default.
What vulnerabilities and incidents actually exist?
Public vulnerability records for ICCP are implementation bugs, not protocol flaws. CISA has published several ICS advisories for specific vendors' ICCP stacks — ICSA-22-249-01 (CVE-2022-38138) against Triangle Microworks' IEC 60870-6/ICCP/TASE.2 library, ICSA-13-240-01 against Triangle MicroWorks' SCADA Data Gateway, and ICSA-22-272-01 / ICSA-23-059-02 against Hitachi Energy's ICCP stack — all denial-of-service or memory-corruption issues in a particular vendor's code rather than a weakness in the IEC 60870-6 specification itself. No publicly documented real-world cyberattack has been attributed to exploiting ICCP specifically: the public technical reports on the 2015 Ukraine grid attack and the CRASHOVERRIDE/Industroyer malware describe other protocols and attack paths, not ICCP. The protocol's documented risk profile is structural — a 30-year-old control-center link with no mandatory encryption — rather than a track record of confirmed breaches through it.
Where does ICCP sit under NERC CIP?
No current NERC CIP standard names ICCP explicitly. CIP-005-7 (Electronic Security Perimeter) and CIP-006-6 (Physical Security) regulate any routable connection generically through External Routable Connectivity and Electronic Access Point requirements, and CIP-012-1 — the standard that specifically governs communications between control centers, ICCP's actual use case — likewise contains no reference to the protocol by name. ICCP is, however, called out explicitly in non-binding NERC guidance, including the CIPC's "Security Guidelines for the Electricity Sector: Control System Electronic Connectivity" and the 2026 CIP Roadmap cited above, both of which treat ICCP links between separate trust domains as needing perimeter protection even though the mandatory standards are written protocol-agnostically.
SoftWSP's OT network assessments passively identify which protocols — ICCP included — are actually carrying control-center traffic across a facility's or utility's network segment, whether Secure ICCP's TLS extension is actually in use, and where unmanaged or unencrypted links create risk, without requiring any agent installed on existing control-center equipment. See OT Snapshot for how that passive read works.
This guide is based on the IEC 60870-6-503 publication page, EPRI's and the national labs' (INL, PNNL, Sandia) own technical reports on ICCP security, CISA's published ICS advisories, NERC's current CIP-005-7/CIP-006-6/CIP-012-1 standard text and its January 2026 CIP Roadmap, all fetched or read directly for this guide. It is not a substitute for a current NERC CIP compliance review and does not constitute a penetration-test report against any specific ICCP deployment.
Sources
- IEC 60870-6-503 (Ed.3, 2014): Telecontrol equipment and systems — Part 6-503: Telecontrol protocols compatible with ISO standards and ITU-T recommendations — TASE.2 Services and protocol
- EPRI Report 1001977: ICCP Protocol — Threats to Data Security and Potential Solutions (2001)
- Pacific Northwest National Laboratory, PNNL-26729 (2017): Cyber Security Analysis of ICCP
- NERC: Critical Infrastructure Protection Roadmap (January 2026)
- CISA ICS Advisory ICSA-22-249-01: Triangle Microworks IEC 60870-6 ICCP/TASE.2 Library (CVE-2022-38138)
- NERC Reliability Standard CIP-012-1: Cyber Security — Communications between Control Centers
Frequently asked questions
What is ICCP/TASE.2, and who standardizes it?
ICCP (Inter-Control Center Communications Protocol) is the common industry name for TASE.2 (Telecontrol Application Service Element 2), standardized internationally by IEC Technical Committee 57 as the IEC 60870-6 series. The core services-and-protocol document is IEC 60870-6-503, with companion parts covering the functional profile (60870-6-702), object models (60870-6-802) and transport profiles (60870-6-601 / TS 60870-6-602). The IEC webstore describes 60870-6-503 as specifying 'a method of exchanging time-critical control centre data through wide-area and local-area networks,' supporting both centralized and distributed architectures. ICCP grew out of EPRI's Utility Communications Architecture (UCA) program in the early 1990s and is used to exchange real-time telemetry, control operations, scheduling and accounting data, and event notifications between utility control centers, balancing authorities, regional transmission organizations and independent system operators.
Does the base ICCP/TASE.2 protocol include authentication or encryption?
No. EPRI's own 2001 assessment of the protocol states plainly: 'ICCP provides access control via bilateral tables maintained on an ICCP server. This is the only data security service provided within ICCP.' Idaho National Laboratory's 2008 common-vulnerabilities report for control systems found that 'unencrypted ICCP communication... may provide many opportunities for an attacker to intercept, understand, and alter the data,' and a 2017 Pacific Northwest National Laboratory report (PNNL-26729) went further, describing ICCP as 'an unprotected protocol' vulnerable to integrity violation, interception or alteration, spoofing and eavesdropping, and noting that 'secure ICCP is rarely implemented if at all' across the industry even though the extension has existed for years.
What does IEC 62351 add for ICCP, and is it actually deployed?
IEC 62351-4, 'Profiles including MMS,' is the part of the IEC 62351 series that specifically defines security profiles for ICCP (IEC 60870-6) and IEC 61850, layering TLS-based transport security (specified generically in IEC 62351-3) and certificate-based mutual authentication on top of the base protocol — an approach the industry calls 'Secure ICCP.' A 2007 Sandia National Laboratories report (SAND2007-3345) documented early Secure ICCP integration work, but a decade later PNNL-26729 (2017) was still describing Secure ICCP deployment as rare. NERC's own January 2026 Critical Infrastructure Protection Roadmap confirms the gap persists at the industry level, stating that 'legacy protocols, such as DNP3, ICCP, and Modbus often traverse unencrypted links that fall outside the current CIP-012 scope.'
Are there known vulnerabilities or incidents tied to ICCP/TASE.2?
Several CISA ICS advisories cover flaws in specific vendors' ICCP stack implementations — for example ICSA-22-249-01 (CVE-2022-38138, Triangle Microworks' IEC 60870-6/ICCP/TASE.2 library), ICSA-13-240-01 (Triangle MicroWorks SCADA Data Gateway) and ICSA-22-272-01 / ICSA-23-059-02 (Hitachi Energy's ICCP stack) — but these are implementation-specific denial-of-service or memory-corruption bugs in individual vendors' code, not flaws in the IEC standard itself. No publicly confirmed real-world cyberattack has been attributed to ICCP exploitation: the publicly documented analyses of the 2015 Ukraine grid attack and the CRASHOVERRIDE/Industroyer malware do not name ICCP as the attack vector. The protocol's main documented risk is structural — the lack of native authentication and encryption described above — rather than a history of confirmed breaches through it.
Does NERC CIP specifically regulate ICCP links?
Not by name in the mandatory standards: a full-text check of the current enforceable versions of CIP-005-7 (Electronic Security Perimeter), CIP-006-6 (Physical Security) and CIP-012-1 (control-center-to-control-center communications protection — the standard most relevant to ICCP's actual use case) shows none of them mention 'ICCP' or 'TASE.2' explicitly; they regulate any routable-protocol connection generically via External Routable Connectivity (ERC) and Electronic Access Points. ICCP is, however, named explicitly in non-binding NERC guidance, including the CIPC's 'Security Guidelines for the Electricity Sector: Control System Electronic Connectivity' and the January 2026 CIP Roadmap cited above, both of which treat ICCP links between trust domains as needing perimeter protection even though no single CIP requirement is written around the protocol's name.
In this cluster


