Asset inventory for inverter fleets: what NERC CIP asks for, and what it does not
What NERC CIP requires in an asset inventory at solar, wind and battery sites, what it does not, and how to build one that scales across a fleet.
A solar, wind or battery portfolio can hold hundreds or thousands of inverters, plus plant controllers, meters, RTUs or RTACs, relays, battery management systems, turbine controllers, firewalls and switches. Owners often assume NERC CIP wants every one of those on a list. At most generation sites it does not. But the sites that skip the inventory usually struggle with the parts of CIP that do apply.
This post separates what the standards ask for from what you need anyway, and suggests how to build an inventory that survives repowering, inverter swaps and firmware updates across a fleet.
Start with the question CIP-002 actually asks
CIP-002-5.1a Requirement R1 asks you to identify BES Cyber Systems by impact level:
- Part 1.1: each high impact BES Cyber System, if any, at each asset.
- Part 1.2: each medium impact BES Cyber System, if any, at each asset.
- Part 1.3: each asset that contains a low impact BES Cyber System, if any.
Notice the change of noun at low impact. For high and medium you identify the systems. For low impact you identify the asset that contains them, which for a generation owner is usually the plant.
Requirement R2 then puts the list on a clock. Part 2.1 asks you to review the R1 identifications, and update them if there are changes, at least once every 15 calendar months. Part 2.2 asks your CIP Senior Manager or delegate to approve them on the same 15-month cycle. Both apply even if R1 identified nothing.
Two checks come before any of this. First, is the plant a BES facility at all? For solar, wind and storage that usually means more than 75 MVA in aggregate, connected at 100 kV or higher. Category 2 inverter-based resources, registering with NERC since May 15, 2026, are not BES facilities and are not under the CIP standards. Registration alone does not bring CIP with it. Second, if the plant is BES, what is its rating? Under Attachment 1, criterion 2.1 rates a generating plant medium impact only at 1,500 MW or more, so almost every individual solar, wind or battery site rates low impact.
What a low impact site is not asked for
CIP-003-9, the standard that covers low impact BES Cyber Systems, says it directly in a note to Requirement R2:
"An inventory, list, or discrete identification of low impact BES Cyber Systems or their BES Cyber Assets is not required. Lists of authorized users are not required."
So an auditor reviewing a low impact plant will not ask for a per-inverter CIP list. The evidence is organised by asset, or by group of assets, the same way the Attachment 1 plan is written.
Where a device list becomes a requirement
That changes at medium impact. Two common routes bring a renewable owner there:
- a single plant at 1,500 MW or more (criterion 2.1); or
- a Control Center or backup Control Center performing Generator Operator functions for 1,500 MW or more in one Interconnection (criterion 2.11). This one catches growing fleets, so re-check it every time capacity is added.
Medium impact systems fall under CIP-010-4. Requirement R1 Part 1.1 asks for a baseline configuration, individually or by group, that includes:
- the operating system or firmware, where no independent operating system exists;
- any commercially available or open-source application software intentionally installed;
- any custom software installed;
- any logical network accessible ports; and
- any security patches applied.
Part 1.2 asks you to authorize and document changes that deviate from the baseline, and Part 1.3 to update the baseline within 30 calendar days of completing a change. You cannot do any of that without knowing, device by device, what is in scope.
Why build a device inventory at a low impact site anyway
The R2 note removes a documentation duty. It does not remove the need to know what is on site. Several parts of the CIP-003-9 plan are hard to meet without a working list:
- Electronic access controls (Attachment 1, Section 3.1). You must permit only necessary inbound and outbound access for routable communication between a low impact BES Cyber System and a Cyber Asset outside the asset. To judge whether each flow is necessary, you need to know which devices talk out of the plant, and to what.
- Vendor electronic remote access (Section 6). You need methods to determine, disable and detect malicious communications on vendor remote access. Equipment makers often reach inverters, battery management systems and turbine controllers through their own paths. If you do not know which devices a vendor can reach, you cannot say how that access is switched off.
- Transient Cyber Assets and Removable Media (Section 5). Knowing which devices technicians connect laptops and USB drives to tells you where the Section 5 checks have to happen.
There is also a reason outside CIP. Executive Order 14421 (August 26, 2026) covers grid-connected inverters and battery energy storage systems among bulk-power system equipment. According to a law-firm summary, DOE must issue implementing rules by December 24, 2026, and may set conditions on equipment that is already installed. Whatever those rules end up saying, an owner who can list make, model and firmware per site will answer faster than one who has to send people to read nameplates.
What to record for each device
A useful inventory does not need to be elaborate. For each device, record:
- site, and location within the site (block, skid, container or turbine);
- device type, manufacturer and model;
- firmware or software version, and the date it was last confirmed;
- network address and the protocols it speaks (for example Modbus TCP or DNP3);
- who manages it: your team, the equipment maker, the O&M contractor or the utility;
- any remote access path to it, and who uses that path;
- whether it sends or receives traffic that leaves the plant; and
- where the record came from (commissioning documents, a network capture, a site walk-down).
The last field matters more than it looks. When a record is wrong, you need to know which source to go back to.
Making it work across a fleet
- Group by template. Renewable plants repeat themselves: the same inverter model and firmware across every block, the same plant controller design across sites. Record the template once and the deviations per site. CIP-010-4 allows baselines by group at medium impact for the same reason.
- Start from what already exists. EPC as-built drawings, commissioning records and SCADA point lists are a good first draft. Treat them as a first draft, because sites change after handover.
- Collect read-only where you can. Many site devices answer on the network. Reading their configuration or watching their traffic is safer than installing software on controllers the maker does not support.
- Tie updates to events. Inverter replacements, firmware campaigns, repowering, a new vendor and a new remote access path should each trigger an update, not wait for the next annual review.
- Tie the review to CIP-002 R2. The 15-month identification review is a natural point to re-confirm the device inventory, and to re-check whether fleet growth has moved a Control Center across criterion 2.11.
Gaps that show up often
- Cellular or other modems installed by an equipment maker for remote support, which may not appear on the owner's network diagrams.
- Devices owned by someone else but inside the fence, such as revenue meters or utility relays. Record who owns them, even if they are not yours to manage.
- Battery containers with their own controllers for the batteries, cooling and fire systems.
- Spare and replacement units that arrive with different firmware than the fleet template.