Two aircraft from the same family, the same type rating, and on one of them a protection the other does not have: the crew discovers it through a difference in the flight manual, rarely through an explanation. The reason is not commercial. The recent variant was processed as a change to the original type certificate, and a three-question line of reasoning decided, domain by domain, whether it stayed on the ancestor’s certification basis or moved to the current amendments. That is the mechanism behind most of the differences observed on the line.
A variant does not open a new file, it amends an old one
A derivative is almost never a new product in administrative terms. The manufacturer does not file an application for a type certificate: it files an application for a change to the existing certificate. The new model is added to the type certificate data sheet, under the same number, with its own line, its own notes and sometimes its own special conditions. The whole regulatory architecture that follows then answers a single question: how far must the body of requirements be brought up in order to process this change, and over what extent of the product?
Two classifications coexist, and they are constantly confused because they borrow neighbouring adjectives. The first sets minor and major against one another: it bears on the effect of the change on weight, balance, structural strength, reliability, operational characteristics or other characteristics affecting airworthiness, and it determines the approval route. The second sets not significant, significant and substantial against one another: it says nothing about the procedure, it says which certification specifications apply. A major change may perfectly well stay on the original basis; a change that looks modest may be significant because it invalidates an assumption.
| Classification | Question asked | What is examined | Consequence |
|---|---|---|---|
| Minor or major | Who approves, and with what involvement of the authority? | Effect on weight, balance, structural strength, reliability, operational characteristics | Approval under the privilege of an approved organisation, or processing by the authority |
| Not significant | Which specifications apply? | General configuration, principles of construction and certification assumptions, all retained | The existing certification basis is carried over |
| Significant | Which specifications apply? | At least one of the three criteria no longer holds | Specifications in force at the date of application, on the change and the affected areas |
| Substantial | Is it still the same product? | Extent of the compliance demonstration to be redone | New type certificate, basis entirely reset |
Three questions, in an order that is not reversed
The sequence is closed: the scope is not discussed before classifying, and nothing is classified before the extreme case has been ruled out.
The first question is that of the substantial change. The criterion is not the apparent size of the change but the extent of the investigation it makes necessary: if the design is modified to the point where a practically complete compliance demonstration would have to be redone, there is no longer a change to be processed, there is a new product. The case is rare, and it is exactly the border between a derivative and a new aircraft that carries a family name for commercial reasons.
The second question, once the substantial case has been ruled out, is that of the significant change. It decides whether the applicable specifications remain those of the existing basis or become those in force at the date of the change application — which may fall several decades after the initial type application.
The third is no longer binary, it carves out. For a significant change, the current specifications apply not to the whole aircraft but to the change and to the affected areas it affects. The rest of the product keeps its original basis. That is where the essence of the file is played out, and it is the only place where the word “negotiation” has any meaning.
When a new version of an aircraft is made, it doesn’t get a brand-new certification. Instead, it’s treated as a change to the original. Three key questions decide if the new version follows the old rules or must meet newer, stricter ones. The most important is whether the change is so big that it’s basically a new plane. If not, the next question is whether the change is significant enough to require updates to some parts of the plane. The final step is figuring out exactly which parts need updates and which can stay as they were.
Consider the Airbus A320 family. The A320neo (new engine option) was a significant change because it introduced new, more efficient engines. However, not all parts of the plane needed to meet the latest regulations. The structure and flight controls had to comply with newer rules, but the cabin and fuel systems could keep the older standards. This way, the A320neo didn’t require a full recertification, just updates in specific areas.
What the three significance criteria actually cover
The general configuration is read geometrically and architecturally: fuselage length and cross-section, wing and engine installation, number and location of the exits, layout of the major systems. A plug added, a pylon moved, a door removed: the configuration is no longer the one that was certified, and neither is the demonstration that relied on it.
The principles of construction concern the materials, the methods of assembly and the load-carrying philosophy. Moving from a riveted metal box to a composite primary structure, from a mechanical control to an electrical one, from one load path to another: the old compliance becomes unusable, not because it was wrong, but because the means of compliance that carried it no longer apply to the new object.
The certification assumptions are the most decisive criterion and the least visible. They are the input data that get forgotten because they appear nowhere on the drawing: maximum number of occupants, mission spectrum, weight and balance envelope, operating environment, assumed failure rates, and above all the crew workload model used in the system safety assessment. A change that adds not one kilogram of metal may be significant if it makes one of those inputs false.
The canonical example is crew alerting. Changing the logic by which failures are presented, the hierarchy of the messages or the split of actions between automation and crew moves no part, but it changes the detection and reaction time assumed in the analyses. The whole chain of demonstration that relied on that time has to be redone. Conversely, a cabin refit that leaves the exits, the occupancy and the routes unchanged may be a major change, heavy to approve, and remain not significant.
The real arbitration: the scope of the affected areas
An area is affected when the change alters its design, its loads, its environment or its operating conditions. Stretching a fuselage affects the structure, the loads, the handling qualities, the takeoff rotation geometry and the evacuation scenario; it does not necessarily affect ice protection or the fuel system. The applicant proposes this scope, with justification, and the authority argues over its edges. A scope that is too wide turns the derivative into a recertification and kills it economically; a scope that is too narrow leaves a new capability demonstrated against old requirements, and it is that case which is paid for in service.

The discussion bears less on the major domains than on the side effects. An increase in thrust obviously affects the engine mounts; does it affect the efflux over the control surfaces, the asymmetric loads after a failure, the sizing of the fin? The answer determines the applicable amendment for requirements that no one had reopened since initial certification.
| Change | What moves | Assumption called into question | Classification tendency |
|---|---|---|---|
| Fuselage stretch | Geometry, loads, weights | Number of occupants and evacuation scenario | Significant, wide scope |
| Re-engining | Nacelles, pylons, systems, performance | Thrust, mount loads, efflux, noise levels | Significant |
| New avionics with a different alerting philosophy | Little hardware, a great deal of logic | Detection time and workload assumed in the safety assessment | Significant despite appearances |
| Increase in maximum weight with no other change | Weight envelope | Load spectrum and damage tolerance objectives | Often significant, depending on the extent |
| Cabin refit on unchanged structure | Seats, partitions, galleys | None, if the exits and the occupancy do not change | Major but not significant |
This table gives tendencies, not verdicts: the classification is pronounced file by file, on the actual design and not on the category of change.
Two ways out, and a floor
Even inside an affected area, the applicant may ask not to apply the most recent amendment. Two grounds are admissible. The first: compliance with the recent text would not contribute significantly to the level of safety of the changed product — the new requirement addresses a mechanism that the existing design handles otherwise, or that has no purchase on it. The second: compliance would be impracticable, in the sense that it would require a redesign unrelated to the purpose of the change.
These two ways out are not boxes to tick: they are demonstrated, point by point, on the requirement concerned and not on a whole chapter. And they always come up against the same floor, the existing basis. The result is a one-way mechanism: a product’s body of requirements rises with each derivative, never the reverse. Nothing, moreover, prevents the applicant from voluntarily complying with amendments more recent than those imposed, which it often does in order to prepare the next variant or to keep design commonality with another programme.
Accumulation, or why the question comes up again with every variant
This is the most counter-intuitive point of the arrangement, and the one that fuels the case against ageing certification bases. Significance is judged against a reference configuration that has been identified, and not against the previous delivery. A long series of independent touch-ups, each beyond reproach taken alone, can take a product very far from what was demonstrated, without any step ever having triggered an upgrade. The reasoning therefore requires looking at the deviation accumulated since the last time the basis was raised in the domain concerned.
The trap. Judging each change against the immediately preceding version is the surest way never to declare anything significant. A fleet can thus drift by several notches without any step having crossed the threshold, until a file review asks which configuration served as the reference — and no one can name it. That reference is documented when the basis is established, not when it is challenged.
What the old basis changes once the aircraft is on the line
An old basis does not mean an aircraft frozen in its era. Recent requirements reach existing fleets through other doors: an airworthiness directive when an unsafe condition is identified, an equipment carriage requirement written into the air operations regulation, an upgrade programme imposed with a deadline. Those doors are not the designer’s: they open at the operator, who bears the timetable and the cost. Hence that familiar observation on the line — two aircraft of the same model, operated under two different authorities, do not display the same equipment even though their certification basis is rigorously the same.
On the crew side, the tangible consequence runs through the operational suitability data. The operational evaluation compares the variant with those already in service and sets the treatment of the differences, from simple information to a simulator session followed by a check. The type rating remains common as long as the differences are handled by training; it splits when the operating assumptions diverge too far. The same reasoning that classified the technical change is thus found again, transposed, in the training decision — and that is logical, since both address the same question: what, in what had been demonstrated, no longer holds?
The right reflex. Before asserting that a function exists “on the type”, go down to the model. The type certificate data sheet states the applicable basis model by model, with its special conditions and its notes; the approved flight manual and the configuration list carry the serial number range concerned. The family name guarantees nothing, the model’s line does.
What to go and check, and the grey area that remains
For certifying staff and the continuing airworthiness management organisation, the practical rule fits in one sentence: an approval of a change applies to a list of models and configurations, not to a family. A supplemental type certificate approved on one variant does not carry over to its neighbour without a demonstration of applicability, even when the airframe looks identical. That is the first discrepancy a file review picks up, and the most expensive to regularise after the fact.
For the training manager, the differences published in the operational suitability data constitute a floor, not a programme. They assume a population of crews already qualified on the reference variant — an assumption that ceases to be true as soon as an operator recruits directly onto the latest model in the family.
For anyone preparing a rating, the differences table reads as a map of the assumptions that have changed. A line on alerting, a line on flight envelope protection, a line on fuel management: each signals a place where the compliance demonstration was redone, and therefore a place where the reflexes acquired on the other variant no longer have quite the same justification.
The grey area, for its part, does not close. Drawing the border between affected and unaffected areas is a matter of engineering judgement, argued and then accepted. Two authorities processing the same derivative do not always place it in the same spot: that is why a variant validated by a second authority sometimes arrives with additional conditions that the state of origin had not judged necessary. Those differences are read in the data sheets and the validation documents, rarely elsewhere, and never in the brochure.
This article describes a mechanism of reasoning, it replaces no source. For a given aircraft, the applicable texts in their version in force, the type certificate data sheet and the operator’s approved documentation prevail over everything above.
Questions fréquentes
Does a major change always impose the most recent amendments?
No, the two classifications are independent. Being major determines the approval route and the level of involvement of the authority; it is being significant that determines the applicable specifications. A major change judged not significant is still processed on the existing certification basis.
How can you tell which certification basis a given aircraft flies on?
The type certificate data sheet is the reference document: it states, model by model, the airworthiness code retained at its amendment level, the special conditions, the equivalent levels of safety and the accepted exemptions. The family name is not enough, and two models on the same data sheet may show different bases.
Is a derivative that keeps the original basis less safe than a new aircraft?
Not mechanically. The frozen basis bears on the design demonstration, but recent requirements reach the fleets by other routes: airworthiness directives, carriage requirements written into the operations regulation, upgrade programmes with a deadline. The point to watch is elsewhere, in the accumulation of changes each judged not significant.
What makes a change substantial rather than significant?
The criterion is the extent of the demonstration to be redone, not the size of the change. If the design is changed to the point where a practically complete compliance investigation would have to be carried out, it is no longer a change but a new product, calling for a separate type certificate. The case is rare and is discussed very early with the authority.
Can you stay below the most recent amendment in an affected area?
Yes, on demonstration, and on two grounds only: compliance with the recent text would not contribute significantly to the level of safety of the changed product, or it would be impracticable. The justification is made requirement by requirement, and the level retained can never fall below the existing certification basis.
Do two variants sharing the same type rating have the same certification basis?
Not necessarily, and this is a frequent confusion. The type rating comes under the operational suitability data and the treatment of operating differences; the certification basis comes under design. Two models may be flown under a common rating while having been demonstrated against amendments separated by several decades.

Leave a Reply
You must be logged in to post a comment.