Indeed beverage grade is another popular terminology. There is a menu of other, more ambiguous, names listed on Linde website.
The US situation with respect to Food Grade / Beverage Grade terminologies appears to have become something of an industrial “hot potato,” notably due to FSMA.
The most recent link/attachments I saw are as of March 2015 / August 2014 –
Food Gases, USFDA, FSMA.pdf 736.12KB
And one retrospective comment - the sample certificate i attached in post 26 demonstrates that "Food Grade" may appear in the Scope (and likely Beverage Grade also). However it remains unclear what compositional elements were required to be classified as FS "hazards" and therefore included in the haccp plan "proper". Whether FSMA's involvement will influence the choices/options for Scope is unclear (to me).
With respect to CO2/haccp/fssc here are a few more related observations -
This document is old (2001) but looks to be an informed explanation of the historical basis of the typical CO2 "contaminant" menus -
I noted this comment -
For each listed parameter, a rationale for its inclusion has been provided. Three rationale definitions were identified:
• Sensory: Any attribute that negatively impacts the taste, appearance, or odor of beverage.
• Process: Any attribute that defines a key parameter in a controlled process and an important consideration in the beverage industry.
• Regulatory: Any attribute whose limit is Set by governing regulatory agencies. Once purified to food grade and to the beverage quality levels described, the carbon dioxide must be handled to avoid contamination. All applicable principles of safety and Good Manufacturing Practices (GMPs) for food additives should be applied throughout the process, from production through delivery, storage and use.
(It is curious IMO that there is no specific mention of Safety (=Regulatory?), some of the currently listed items, eg HCN, benzene surely qualify)(benzene would fall in the 2001 "aromatic hydrocarbons" category though)
(Note that ~50% of  items are sensory, if still unchanged these would not conventionally be of haccp interest)("Process" is ambiguous)
From a FS POV, there seems to have been a similarity of interpretive approach between the "food gases" and compressed air (CA) with respect to "specifications".
For interest/comparison, can see the CA consequences in this relatively recent Standard for CA -
cdx1 - BCAS Food Grade Compressed Air - COP,Rev.001,2006.pdf 320.28KB
During the production, processing, handling, packaging and transporting of food at all stages from source to consumer, the safety of the food may become compromised and then subject to a recall. One source of possible contamination is when compressed air is used in conjunction with the food chain.
Contaminants commonly associated with compressed air are categorised in the form of dirt (solid particles), water (as humidity) and oil. Other influences are the presence of microbiological organisms such as bacteria which may be transported through a compressed air system or propagated due to the conditions existing in a compressed air system.
Contaminants that may be a potential hazard in food for human consumption need to be subject to control as a lack of control could result in a potential prosecution. Compressed air may come into direct or indirect contact with food. Where compressed air comes into direct contact with food, in particular during production or processing, this requires a much higher level of contaminant control and particular attention needs to be placed on contaminants added during the compression and the distribution process. e.g. bread packing, fluidised bed in the transfer of flour from a tanker etc.
Guidance on the Hazard Analysis and Critical Control Points identifying the elements needed to meet the users' health and hygiene requirements for compressed air systems are included as identified in The Food Hygiene (England) Regulations 2005, Schedule 2.It identifies contributory elements associated with food safety standards as identified in BRC requirements.
The HACCP process should be applied to determine the level of contamination by the use of the tests identified in clause 7 and then to verify if these levels are within the values identified in clause 6. If the values are outside of those in clause 6 then remedial action shall be taken.
+ sections 6,7
Note that the precise correlation between "contaminants" and "safety" is somewhat "implicit". Also note the word "prosecution" in above quote, ie a Regulatory factor.
(such ambiguities have created considerable, only partially resolved, confusion in the haccp handling of CA, notably within the SQF sub-forum)
Do any of the other FS Standards accept CO2 for FS Certification purposes, or discuss its spec. requirements as an ingredient ? A possible reference source if yes.
i also noted the FSSC viewpoint on food gases -
cdx2 - fssc scope of food gases.png 131.25KB
I am not aware if Codex defines a purity specification for CO2 as a food additive. If so this could be the controlling “contaminant menu” from FSSC's POV . If not, I daresay the choice is open. Regardless, any local food regulations will presumably supercede Codex if they exist, such as EC, JEFCA, FDA. Possibly even Beverage Grade.
Not my area but i noticed the CO2 process extract attached below. I deduce that, at least in principle, achievement of the various contaminant spec. menus may simply be a case of appropriate filtration. if so, the lengthiest list of requirement(s) would seem not a major hurdle unless due economic limitations. I doubt FSSC will care if you exceed their minimum expectations.
cdx3 - carbon dioxide purification.png 996.36KB
(It, again, implies that several of the listed impurities are related to flavour rather than toxicity. Emphasizes the necessity of well-understanding yr product's intended application)
Seems to me that, based on own specific interest, the OP has 1 initial query to address to the proposed FSSC auditor –
What “Grades” are acceptable to be referenced within the Certificate’s Scope ?
If the answer is that it is totally determined by presented, validatable, data then I guess all options are available.
Afaik, “sensory” factors are not required to be considered within the fssc hazard analysis scheme but could be included within the overall FS system via a PRP (ie iso22002-1).
@ alvindman - hope the above of some interest/assistance. It's an interesting topic
PS - I'm surprised ISBT made no mention of US Regulatory bodies, eg FDA. Perhaps because a US-centric scenario only. So far anyway.
PPS - JFI - some of content in this thread overlaps a parallel thread -