Module 2: Security-level allocation(2/3)

Deriving SL-T from risk analysis results

40 min4 min readRef: IEC 62443-3-2 §6

title: "Deriving SL-T from risk analysis results" duration: "40 min"

Connecting risk scores to Security Levels

You now have:

  • A risk matrix score (Low / Medium / High / Critical) for each zone × threat scenario.
  • A tolerable risk threshold defined by management.
  • An understanding of the SL vector and what each level means.

This lesson teaches the systematic process of translating risk scores into SL-T assignments.

The derivation process

Diagram

Step 1: Map risk score to threat-actor capability

The risk score tells you which threat actors are credible for each zone. Higher risk means you must defend against more capable adversaries:

Risk scoreCredible threat actorsMinimum overall SL-T
LowAccidental, casualSL 1
MediumMotivated individual, hacktivistSL 2
HighSophisticated group, ransomware crewSL 2–3
CriticalNation-state APTSL 3–4

Step 2: Determine SL per FR

For each FR, consider which specific threats apply:

Worked example

Zone 1 (Process Control), rated High risk:
FR 1 (IAC): The primary threat is unauthenticated access to PLCs. Modbus and S7comm have no native auth. Compensating control: protocol-aware firewall with per-source ACL. Required: SL 2 (unique source identification).

FR 5 (RDF): The primary threat is lateral movement from the supervisory zone. Required: SL 3 (protocol-aware deep-packet inspection on zone boundary).

Some FRs may be higher or lower than the overall SL-T. This is normal — the vector allows precision.

Step 3: Assemble the SL-T vector

Combine the per-FR values into the vector:

ZoneFR 1FR 2FR 3FR 4FR 5FR 6FR 7Overall
Zone 1 (Control)2221322SL 1
Zone 2 (SIS)3332333SL 2
Zone 3 (Supervisory)2221222SL 1
Zone 4 (Site Ops)2222222SL 2
Zone 5 (Enterprise)1111111SL 1

Key takeaway

Notice the gap

Zone 1 has FR 5 at SL 3 but FR 4 at SL 1. The overall zone SL is pulled down to SL 1 by the weakest element. If the asset owner wants Zone 1 at SL 2, they must raise FR 4 from SL 1 to SL 2 — which means implementing SR 4.1 (information confidentiality) at the enhanced level.

Step 4: Validate against ALARP

For each zone, check:

  1. Is the assigned SL-T sufficient to reduce the risk below the tolerable threshold?
  2. Is the cost of achieving the SL-T proportionate to the risk reduction?
  3. Are there any FRs where the SL could be lowered without increasing risk above the threshold?

If the SL-T is insufficient, increase it. If the cost is grossly disproportionate, document the justification for accepting a lower SL.

Worked example: SIS zone

Risk score: Critical (Consequence 4 × Likelihood 3). Tolerable risk: All Critical risks must be reduced to High or lower. Required SL-T: SL 3 (minimum to defend against sophisticated groups with ICS expertise).

FR 4 (DC) challenge: Encrypting TriStation traffic is not supported by the current Triconex firmware. Option: upgrade firmware (vendor cost: $45K, 2-day outage). Alternative: accept FR 4 at SL 2 with compensating control (dedicated VLAN, no other traffic on the segment).

Decision: Accept FR 4 at SL 2 with compensating controls. Overall zone SL-T remains SL 2, but the compensating controls reduce the residual risk for FR 4 to within ALARP. Document the decision in the risk register.

Common pitfalls

  • Over-assigning SL 4 — SL 4 requires hardware-backed credentials and formal verification. Very few commercial ICS products are certified to SL 4. Do not assign it unless the threat model genuinely includes a nation-state APT with unlimited resources.
  • Under-assigning FR 5 — network segmentation is the highest-leverage control. If your risk analysis shows lateral movement as a primary threat, FR 5 should be at least SL 2 in every zone.
  • Ignoring the vector — an overall SL-T of "SL 2" without the per-FR breakdown is useless. The vector tells you exactly which controls to implement.

Key Takeaways

  1. Map risk scores to threat-actor capability to determine the minimum overall SL-T.
  2. Determine the SL per FR based on the specific threats to each zone.
  3. Assemble the full SL-T vector — the per-FR breakdown drives control selection.
  4. Validate against ALARP — ensure the SL-T is proportionate to the risk and cost.
  5. Document any compensating controls for FRs that cannot reach the target level.

Knowledge Check

3 questions — test your understanding before moving on.

  1. Q1.What is the first step when deriving SL-T from risk analysis results?

    • Choose the most expensive control available.
    • Map the risk score to the credible threat-actor capability to determine the minimum overall SL-T.
    • Assign SL 4 to every zone as a precaution.
    • Count the number of CVEs per device.

    The derivation starts by mapping the zone's risk score to the threat-actor capability class. Higher risk (Critical) means defending against more capable actors (nation-state, SL 3–4). Lower risk (Low) means basic controls suffice (SL 1).

  2. Q2.Why is it important to specify the SL-T as a per-FR vector rather than a single number?

    • Because regulators require exactly seven data points.
    • Because different FRs may require different SL levels depending on the specific threats to the zone — the vector drives precise control selection.
    • Because a single number is not supported by the IEC 62443 standard.
    • Because each FR corresponds to a different budget category.

    Different threats affect different FRs. A zone may need SL 3 for FR 5 (network segmentation) because lateral movement is the primary threat, but only SL 1 for FR 4 (data confidentiality) because the data is not sensitive. The per-FR vector tells you exactly which controls to implement.

  3. Q3.When should you avoid assigning SL 4 to a zone?

    • When the zone contains safety-critical assets.
    • When the threat model does not include nation-state APTs with unlimited resources, since very few commercial products are certified to SL 4.
    • When the zone is connected to the internet.
    • When the zone contains more than 10 devices.

    SL 4 requires hardware-backed credentials, formal verification, and continuous automated response — capabilities that very few commercial ICS products support. Assigning SL 4 without a genuine nation-state threat model inflates costs without proportionate risk reduction.