Worcester EA229 Fault Code: Why the Flame Cuts Out

What Worcester EA229 means

On the Greenstar CDi Compact ErP models referenced here, EA229 means a flame signal has been lost during operation. The boiler had reached a different stage from an EA227 startup failure. The code identifies that interrupted operating condition; it does not identify the part or external condition responsible.

Confirm the exact model and full displayed message. Worcester's newer Greenstar 1000 manual also describes numeric 229 as flame failure during burner operation, but that does not mean every generation displays the same EA prefix or uses identical controls. The manual for your own appliance remains the reference.

A reset may restore service temporarily, but repeated loss of the flame signal needs investigation. Do not run the boiler with a trap removed, alter gas components or repeatedly restart it to keep the house warm. Normal operation returning for an hour does not establish a repair.

If you smell gas, leave the building and call 0800 111 999 from outside. Stop using an appliance that appears unsafe or develops abnormal banging or other concerning symptoms. For a routine lockout, the useful starting information is when the fault occurred during normal use, which service was running and what work has already been done.

Sources: Worcester Greenstar CDi Compact ErP manual, EA229 fault table · Worcester Greenstar 1000 manual, numeric 229 · HSE domestic gas safety

The difference between losing a flame and failing to start

EA227 and EA229 can both leave a home without heating, but they describe different points in the sequence. The first concerns unsuccessful flame confirmation after ignition. The second concerns loss of a flame signal during operation. That difference makes the timing of the event useful to the investigation.

A household may notice that the boiler starts, provides some heat and then stops. That fits the practical question raised by an operating-stage fault, but it is still not a component diagnosis. The controls report what they detect; only appropriate testing can establish why that condition arose.

Do not try to observe a flame through an opened casing or time internal components yourself. The normal screen, any existing fault history from previous visits and a description of ordinary operation are enough for the initial enquiry. An engineer can decide which controlled tests are needed.

If several codes have appeared, retain their order rather than replacing all of them with the phrase ignition fault. A record of EA229 followed by another message answers a different question from a single photograph taken after several resets. The aim is to preserve the sequence, not to provoke new failures to complete it.

Why a boiler can work for hours before the code returns

An intermittent fault creates a difficult practical situation: the boiler may work during an appointment yet fail again later. A period of successful operation establishes that it could run under those conditions. It does not prove that every condition associated with the reported failure has been reproduced.

Think about the ordinary circumstances of the event. Did it occur during heating, hot-water use or a change between demands? Was the boiler already running for some time? Was there a notable weather change? These are observations to share, not instructions to design a household stress test.

The engineer may need to distinguish a fault that happens at a particular operating point from one that is present all the time. That distinction affects what can reasonably be concluded from a short visit. It is fair to ask whether the original symptom was reproduced and what uncertainty remains if it was not.

Avoid interpreting every gap between failures as proof that the most recent adjustment worked. If the original fault was intermittent, a quiet afternoon may simply be an interval without the triggering conditions. A repair should be assessed against the evidence supporting it and the relevant operating checks, not only the number of hours until the next message.

A recent homeowner account explains why repair history matters

In a December 2025 DIYUK discussion, a homeowner described repeated EA229 after earlier fan work and later condensate-trap work. Resets restored operation for a time, then the fault returned. The owner described an internal drain route and questioned what else could explain the recurring shutdown. No verified final resolution was available.

The important point is not that either replacement was necessarily wrong. The public account does not provide the measurements or complete reports needed to judge that. It shows why the next investigation needs the history of what changed, what result followed and how long the improvement lasted.

Bring those records together rather than starting again with a bare code. A list of replaced parts can be useful, but the findings supporting each replacement are more valuable. They help distinguish a previously confirmed defect, an exploratory repair and an assumption that may need reconsideration.

The discussion also contained unsafe suggestions about operating the boiler with drainage components removed. We do not reproduce that procedure. A public thread can reveal a customer's unresolved question without providing a safe method for answering it. The appropriate next step is controlled investigation by someone competent for the appliance, using the full history.

Source: Public homeowner account: recurrent EA229 after earlier repairs

Drainage may be relevant even when nothing outside is frozen

The CDi Compact ErP fault guidance identifies condensate-related conditions among possible contexts for EA229. That is a reason for appropriate investigation, not proof that a trap needs replacing. The drain route, appliance-side components and wider conditions should not be collapsed into one interchangeable diagnosis.

An indoor route changes the question about exposure to freezing temperatures. It does not, by itself, establish that all drainage is satisfactory. Conversely, a visible outdoor section during cold weather does not prove that freezing caused the current code. Route and condition are separate questions.

Tell the engineer what is safely visible and where the pipe disappears. If previous drainage work was done, provide the report and whether the symptom changed afterwards. Those details are more useful than insisting that a route must be clear because it is indoors or must be blocked because the weather is cold.

Do not disconnect a hose, remove the trap or leave a drain component off while running the boiler. Nor should a household attempt to prove a diagnosis by recreating someone else's forum experiment. If access to concealed drainage is proposed, agree what evidence supports that location and how the investigation will distinguish a route problem from an appliance problem.

Weather observations should inform testing, not decide the cause

The CDi Compact manual also identifies strong wind at the flue as one possible context for loss of flame signal. This does not make every windy-day shutdown a flue fault, and it does not authorise a homeowner to alter a terminal, add a cover or obstruct the opening.

A useful weather observation is specific: the fault occurred during a particular storm or has repeatedly appeared in similar ordinary conditions. An isolated coincidence is weaker evidence than a consistent history, but neither replaces inspection and testing. State what happened without converting it into a guaranteed explanation.

Do not climb to the flue or reach out of a window to inspect it. If something unusual is visible safely from ground level, describe it. If the route or terminal is inaccessible, that is an access issue for the person assessing the installation, not a reason to improvise a ladder inspection.

Keep nearby building work in the history if it seems relevant, but avoid attributing the fault to a neighbour or contractor without evidence. The appropriate question is whether anything affecting the installation needs assessment. A measured explanation is more useful than a confident accusation based on the first windy day after work nearby.

Supply and flame-sensing checks are not homeowner experiments

The newer Greenstar 1000 installer guidance includes professional checks of gas conditions, combustion-related settings and flame-detection connections for its numeric 229 fault. These categories explain why an operating-stage shutdown can require more than a reset. They are not a sequence for the homeowner to perform.

A functioning hob or a positive account balance does not establish the conditions at the boiler while it is running. Report known supply information, but do not light another appliance solely to test gas availability or manipulate a valve to see whether performance changes.

Similarly, the absence of a visible problem outside the casing does not establish that detection is reliable. The household cannot assess an electrode connection or electrical signal by listening to the boiler. That is why a code describing flame loss should not become an instruction to clean or replace an internal item yourself.

Ask the engineer to explain the finding that supports the proposed repair. If the recommendation concerns supply, detection or another operating condition, it should connect to evidence from the appliance and installation. The explanation need not include a hazardous test procedure to be clear enough for the customer to make an informed decision.

Hot water restored but heating still absent is a separate outcome

A December 2022 PistonHeads account described a Greenstar 30Si Compact with EA229. The owner reported that subsequent gas-valve work restored hot water but not central heating, with further attendance planned. The final resolution was not shown. That account does not establish blame or prove which remaining component was at fault.

It does show why the result should be checked against each service the household lost. A combi providing hot water at a tap is one useful outcome. It does not establish that the heating controls and connected system are also providing the required service.

If one function works after repair and another does not, describe that accurately. It may be a remaining issue within the original investigation or a separate problem that now needs assessment. Do not assume that the same displayed code must still explain everything, especially if the code has cleared but the service has not returned.

A handover should therefore distinguish appliance fault clearance from the actual heating and hot-water results. Ask what was tested and what was outside the agreed scope. That avoids a conversation in which the engineer means the burner operated while the customer understandably hears that the whole household heating problem is solved.

Source: Public homeowner account: hot water restored but heating still absent

Controls can complicate the history without being the proven cause

Smart schedules, room controls and hot-water settings can change when demand occurs. They can therefore affect the circumstances in which a fault becomes noticeable. That is different from proving that a smart thermostat caused loss of the flame signal.

The December 2025 homeowner account mentioned trying to understand the timing around the heating controls. The unresolved question was whether the pattern explained the recurring fault. Without a verified diagnosis, it would be wrong to turn that concern into advice to replace or rewire the controls.

Use the normal user interface to describe the intended schedule and whether a demand was present. Do not bridge terminals, force a valve or alter wiring to bypass the control system. An engineer can decide whether the control context is relevant and what needs checking safely.

If settings were changed while troubleshooting, include that history. A fault that first appeared before a schedule change should not later be remembered as having started because of it. Separating the order of events makes the evidence more useful and reduces the risk of replacing an unrelated control simply because it was the last thing someone adjusted.

Reset once if permitted; do not build a daily workaround

Follow the exact user manual's normal reset instruction where the appliance appears safe and there is no contrary instruction. If the code remains or returns, stop repeated attempts and arrange investigation. Reset procedures differ, so a generic press-and-hold duration is not appropriate across Worcester generations.

A daily reset routine can conceal how often the appliance is failing. The household may begin to describe the boiler as working because heat eventually arrives, while an operating-stage fault is recurring repeatedly. That distinction matters when deciding whether the underlying issue has been resolved.

If the reset restores operation for only a short period, tell the contractor that interval and the ordinary use preceding recurrence. Do not run repeated hot taps or heating cycles solely to force the fault back. An existing pattern is enough to justify investigation without asking the household to perform a test programme.

Stop using the appliance if abnormal noises, gas concerns or other unsafe signs accompany the fault. A successful reset does not cancel those concerns. The purpose of the control is to request another normal attempt where appropriate, not to certify the installation or override the need for assessment.

Agree a repair process suited to an intermittent fault

An intermittent EA229 investigation benefits from continuity. Provide the exact model, full code, previous reports and the practical pattern in one place. Ask whether the appointment is for fault investigation, routine servicing or a follow-up to earlier work, because those descriptions can carry different expectations.

Before authorising a replacement, ask what finding supports it and how success will be assessed. If the original symptom could not be reproduced, that limitation should be explicit. It is possible to recommend a justified next step without pretending that uncertainty has disappeared.

For AK enquiries in Greater London, include the postcode and whether heating, hot water or both are affected. AK offers a 24/7 phone and emergency dispatch service, but parts availability and the investigation required depend on the appliance. Attendance is not a promise of a same-visit repair.

The £60 attendance charge covers arrival only; agree boiler investigation and repair scope separately. The published plumbing hourly rate is not a universal rate for every type of gas-boiler work. Ask about any parts, follow-up attendance and remaining uncertainty before approving the next stage, and keep the resulting findings for any subsequent visit.