This is an interesting question that has a less than satisfactory answer. When brewers discuss carbon dioxide purity the elephants in the room often include oxygen, oil and sulfur compounds. While all of these compounds may have no effect on food safety (oils may if they are derived from petroleum distillates, for example) they do all have the potential to damage beer. So in practical terms there is a very real difference between food-grade and high purity carbon dioxide.
The fact is that all companies I know who sell carbon dioxide carry one stock gas and sell this gas to various customers who in turn use the gas in a wide variety of applications, brewing and soda dispensing being just two of many. When it comes to food grade, the primary difference lies not in the gas, but the gas container. If liquid carbon dioxide is transported from a gas supplier to a brewery that in turn adds the gas to beer, the supply chain from the gas company to beer must all be clean, sanitary and compliant with standards that are consistent with “food grade.” If this last clause seems ambiguous, it is. Since local health codes differ across the world, and often times within the confines of small regions, there is not a great definition of food grade.
In the broadest sense of the term, food grade simply means that a product is processed, handled, stored, transported and used in a manner that does not permit spoilage or the introduction of substances that may be injurious to the consumer. So let’s look at this from a practical point of view. Bottles used to transport compressed/liquid gases can potentially be used for multiple purposes. A bottle used to transport liquid propane could subsequently be used to transport carbon dioxide. This practice is not acceptable because the residual propane would contaminate the carbon dioxide. One very important thing that is done to prevent this sort of contamination is defining what types of gases are filled into bottles.
Another thing that goes along with handling gases intended for human consumption is cleaning of containers and the segregation of tools, fittings and hoses that may serve as vectors. Again, this is all fairly approximate from what I have seen. When it comes right down to it, I really doubt there is much oversight by US health inspectors when it comes to carbon dioxide because this ingredient does not have a history of causing health problems and does not draw the attention from regulators. In my opinion, this is a good thing.
But your question is really about brewing and we still have the elephants lurking in the corner. Let’s start with oxygen. Oxygen in ubiquitous on planet earth, is detrimental to beer and is almost always present in carbon dioxide gas at some level. Typical levels are between 0.05–0.5% and this amount of oxygen is sufficient to oxidize beer. This is a very broad topic and I will resist commenting more about oxygen and carbon dioxide. The important thing is for brewers to recognize this fact. Carbon dioxide can also contain sulfur compounds because of some of the sources that gas suppliers use for carbon dioxide. Fuel ethanol, fertilizer and petroleum processing all yield carbon dioxide as a byproduct and most of the carbon dioxide gas that we buy originates from these sources, and all of these sources can contain sulfur contaminants that brewers do not want in beer. And finally, there are oils from compressors that may be present in carbon dioxide.
Larger breweries treat carbon dioxide like any other ingredient and carefully monitor carbon dioxide purity for a number of obvious reasons. As with water, carbon dioxide is often filtered to remove flavor-active compounds. This is often performed as a prophylactic measure to simply guard against the possibility of stinky gas causing a problem. A common place to use these filters is between carbon dioxide bottles and beer kegs tapped in bars and restaurants.
I recently attended the 2014 Brewing Summit in Chicago, a joint meeting of the Master Brewers Association of the Americas (MBAA) and the American Society of Brewing Chemists (ASBC). During a break between meetings I had a great conversation with a fellow brewer about some of the hoopla stirred up by bloggers regarding ingredients in food and beer. My friend made a great point about carbon dioxide. In a nutshell, his point was that many brewers have a lackadaisical approach to carbon dioxide. For example, when asked to talk about the “four ingredients” of beer most brewers could easily gnaw the ear of most beer consumers. But the same brewers often times do not know anything of substance about their carbon dioxide supply. My friend’s point was that the carbon dioxide supply question is very easy to answer when beer is naturally carbonated during tank or bottle conditioning.
Thank you for asking this question. I feel like a captain who just felt a little bump on the bow and peered into the water to see a great berg lying beneath the surface. This topic is one that is infrequently discussed among homebrewers and small craft brewers, and it also lacks any real coverage in most of the books, magazine articles and blog posts written for this group of brewers.