Energy Networks Association ER G99 superseded G59 in 2018 and has been quietly tightened ever since. A 1 MW BESS connection landing on a UK DNO's 11 kV network now needs LFSM-O, LFSM-U and frequency-vector-shift evidence that no first-time connector tends to have ready. We walk through what a clean application package actually contains.
This isn't just bad luck; it’s a recurring scene in containment offices and engineering consultancies across Great Britain. The transition from the simpler G59/G83 world to the comprehensive G99 framework, which embeds the sprawling European ‘Requirements for Generators’ (RfG) network code into UK practice, has been anything but smooth. A G99 application isn’t a form-filling exercise. It is a technical negotiation, a detailed engineering prospectus for a power-generating asset that must convince a deeply skeptical network owner you won’t cause them any trouble. And all too often, these applications fail to persuade.
The Type A-D Identity Crisis
The very first box-checking exercise is often the first place an application goes wrong. Your Power Generating Module’s (PGM) classification under G99 dictates the entire subsequent process. This isn’t just about the registered capacity in megawatts; it’s a matrix of capacity and connection voltage. Get this wrong, and every subsequent submission is built on a faulty foundation.
The common mistake? Oversimplification. An engineer sees a 950 kWp solar farm and instinctively thinks "Type A," assuming the sub-1 MW threshold is the only criterion. But if that site is connecting at 11 kV, G99 is crystal clear: it’s a Type B PGM. The rules for Type A PGMs only apply to those connected at a voltage below 110 kV with a capacity of less than 1 MW *and* connected at a voltage below 1000V AC. That 11 kV connection bumps you into a whole new world of requirements.
Why does it matter? The obligations escalate sharply:
- Type A: Simple compliance verification. Mostly just requires manufacturer-certified and tested equipment.
- Type B: Introduce the pain of Fault Ride Through (FRT), dynamic voltage support, and the need for more detailed modelling data.
- Type C & D: We’re into full-blown transmission-level compliance, requiring sophisticated, validated simulation models (like a PSSE or PSCAD model), detailed protection studies, and potential involvement from National Grid ESO, even if you’re just connecting to a DNO’s network.
Declaring a Type B PGM as a Type A is an instant rejection. It signals to the DNO that you haven’t grasped the fundamentals, and they will rightly assume the rest of your submission is equally flawed.
Fault Ride Through Fantasies
No single area causes more G99 headaches than demonstrating FRT capability. The DNO’s primary concern is system stability. They need absolute certainty that your generating plant will not trip off during a minor voltage dip on the wider network, which could worsen the initial fault. They need your plant to "ride through" the problem.
Here’s where applications fall apart:
1. The Generic Datasheet: Submitting a glossy manufacturer’s brochure with a vague claim of "G99 Compliant" is the cardinal sin. The DNO does not care about the marketing material. They need the specific voltage-against-time FRT envelope for the exact inverter or generator model you are using, demonstrating its ability to stay connected during the voltage dips defined in G99 section 10.6.
2. Ignoring the Point of Connection: A manufacturer might test their inverter and prove FRT capability at its own AC terminals. This is useless to the DNO. G99 requires FRT to be demonstrated at the Point of Common Coupling (PCC). The impedance of your site’s transformers and cables will cause an additional voltage drop during a fault, meaning the voltage at your generator’s terminals will be lower than at the PCC. Your simulations must account for this by modelling the fault at a "zero electrical impedance" point on the grid side of the PCC.
3. Active vs. Reactive Current Injection: Surviving the voltage dip is only half the battle. G99, aligning with the EU code, requires the PGM to inject reactive current to support the grid voltage during the fault. Your application must clearly state the reactive current injection strategy and prove via modelling that your equipment can deliver it. This is a dynamic response, and a simple static datasheet cannot prove it. You need a simulation report from a trusted power systems analysis package.
Submitting a proper FRT study is non-negotiable. If your chosen inverter manufacturer cannot provide the detailed simulation models (often in PSCAD or PSS/E format) and type test reports, you have chosen the wrong inverter. It’s that simple.
Modelling and Simulation Missteps
A G99 submission is accompanied by a shadow document: the simulation model. This is your plant’s digital twin, and the DNO’s engineers will interrogate it mercilessly to predict its behaviour. Inconsistencies between the model and the application are a fast track to rejection.
The most common modelling errors are maddeningly basic.
- Parameter Mismatches: The application’s protection schedule states the G59 relay is set to trip on a 2-second over-voltage, but the model’s relay parameters show a 1.5-second delay. Rejected.
- Incorrect Short Circuit Contribution: The form states the PGM’s maximum fault contribution is 1.8 MVA, but the model simulation shows it pushing 2.1 MVA into a test fault. Rejected.
- Outdated or Unvalidated Models: For Type C and D sites, the DNO (and often National Grid) will require a dynamic model in a very specific format. Providing an unencrypted, unlocked, and validated model that can be integrated into their network-wide simulations is key. The new Engineering Recommendation G98 process for model pre-validation is a crucial step that many developers miss, leading to delays. If you're working on a larger project, getting your head around the package substations that can streamline this process is a worthwhile investment.
Your model is your testimony. If it has holes, your entire case falls apart. The DNO is not your debugger; they will not fix your model for you. They will simply send it back.
The Devil in the Details
Beyond the big-ticket items of FRT and modelling, a surprising number of applications are sunk by what appear to be minor administrative and technical errors. To a DNO, however, these errors are red flags that point to a lack of diligence.
- Ambiguous Single Line Diagrams (SLD): The SLD is the master document. Common failures include not clearly marking the PCC, failing to specify cable lengths and impedances, or showing protection zones that don’t align with the proposed relay settings.
- Vague Protection Philosophy: Simply stating "G59/G99 protection will be installed" is insufficient. A protection narrative and settings schedule must be provided, demonstrating coordination with the DNO’s existing upstream devices. You must prove your plant won’t trip before the DNO’s fuse blows, and vice versa. It’s a delicate dance that requires specific data exchange.
- Outdated Application Forms: The Energy Networks Association (ENA) updates G99 and its associated forms with unnerving frequency. Submitting an application on a form from a previous amendment is the easiest way to get it rejected before an engineer even looks at it. Always download the latest forms directly from the ENA portal.
- No Commissioning Plan: An application is a promise. The commissioning test sheet is the proof. Your submission should include a G99-compliant commissioning plan detailing how you will test and verify every parameter promised in the application, from reactive power capability to FRT performance. This shows the DNO you’re thinking about the full project lifecycle.
These may seem like tedious details, but in the highly regulated, risk-averse world of network connections, they are the very substance of a successful application. Fail here, and you telegraph incompetence. If navigating these complexities feels overwhelming, our expert teams can guide you through the entire process; just get in touch before you hit submit.
Key Takeaways
- PGM classification is everything. Misunderstanding the voltage and capacity thresholds for Type A/B/C/D PGMs is a foundational error that invalidates your entire submission.
- FRT evidence must be specific and simulated. Generic manufacturer datasheets are worthless. You must provide a simulation study proving FRT capability at the Point of Common Coupling, not the generator terminal.
- Your model is your bond. Any inconsistency between your application forms and your power system simulation model—no matter how small—will be found and will trigger a rejection.
The Engineer's Takeaway
A G99 application is not a bureaucratic hurdle to be cleared; it’s the primary engineering document justifying your asset
