I'll start by giving credit to my sources. Most of my cap and contract information comes from Spotrac, but the estimates for the price of RFA tenders come from OTC. Information on the prices of veterans' minimum salaries (which
determine ERFA tenders) can be found on p. 146 of the CBA of 2011. In order to simplify the calculations I make, I will round numbers to the nearest $1000.

First order of business, explaining the top 51 rule for those who need the explanation:

During the offseason, all 32 teams are building up to their training camp rosters of 90, but only the top 51 out of those 90 cap hits are counted against the cap. After League mandated roster cuts to 53 occur following TC and preseason games, all 53 regular roster player cap hits, plus cap hits for players on IR, PUP, and practice squad will count, and must fit under the cap. The Broncos currently have more than 51 players under contract for 2019, so every time a player with a higher cap hit is signed to the roster, a lower cap hit will fall out of the top 51. Conversely, every time a player with a higher cap hit is cut or traded, a lower cap hit will climb into the top 51. In order to simplify keeping track of who is in or out of the top 51, I will make all my adjustments for the top 51 rule near the end of the post with one exception which will be in my next paragraph.

Establishing my starting point:

Spotrac currently (at 3:10 PM Central Time on February 18, 2019) shows the Broncos with $22.735M in top 51 cap space. That number assumes a 2019 League cap increase to $189M, includes $8M in rollover of unused 2018 cap space, accounts for dead money to date, accounts for Brandon Marshall leaving in free agency, and has both Case Keenum and Joe Flacco on the roster and in the top 51. I'm going to make 2 relatively minor adjustments to that number. The first involves a former player named Jimmy Williams. He was a CFA last year who did not report to camp. Spotrac shows him on the roster with a $570K cap hit. I'm going to add that $570K to the current Spotrac number, and then apply the top 51 rule by subtracting the $495K cap hit for Horace Richardson which is my new #51. That gives me $22.81M. Now I'm going to subtract $677K for player payments for OTAs. (I calculated that number by multiplying $235 times 2880. Those numbers come from pages 90 and 133 of the CBA of 2011). That gives me an adjusted $22.133M in top 51 cap space.

Cutting Keenum and Darian Stewart:

I'm going to assume Keenum gets cut, and signs elsewhere for a veteran's minimum salary of $805K. That results in a cap savings of $11.805M after applying the offset. I'll add that to my $22.133M giving me $33.938M. I'm also assuming Darian Stewart gets cut resulting in a cap savings of $3.569M. Adding that gives me $37.507M. But for now I'm only counting 49 cap hits. I'll account for that later.

Accounting for RFA tenders:

Spotrac shows the Broncos with 5 RFAs: Shelby Harris, Deiontrez Mount, Casey Kreiter, Kevin Hogan, and Jordan Taylor. OTC projects prices of $4.429M for a 1st round tender, $3.11M for a 2nd round tender, and $2.035M for an original round/right of first refusal tender. Only 2 of these players are important to me, Shelby Harris and Casey Kreiter, but for now I'm going to assume Mount is allowed to walk away in free agency, and Hogan (Flacco needs a backup) and Taylor (we're a little thin at WR, and he's a sure handed if unspectacular punt returner) get tendered at the original round level. RFA tenders are only guaranteed for injury, so we won't lose anything if a tendered RFA ends up getting cut at the end of preseason. I'm not going to tender Casey Kreiter; I'm going to extend him since top paid veteran long snappers only make a little over $1M average per year. I'll assume that results in a nice round $1M cap hit for him for 2019. I'm going to tender Shelby Harris at the 2nd round level. $3.11M for Shelby Harris + $1M for Kreiter + $2.035M for Hogan + $2.035M for Taylor comes to $8.18M. I subtract that from the $37.507M I ended up with in the paragraph above, and I have $29.327M left. But that's NOT top 51 cap space. I had 49 players I was counting, I added 4, so I'm now counting 53 players against the cap.

Accounting for ERFA tenders:

ERFAs receive non-guaranteed one year tenders for veteran minimum salaries. So I think I'll just tender all 7 of them. Jerrol Garcia Williams and Tim Patrick have 1 credited season, so require tenders of $570K each. Elijah Wilkinson, Joe Jones, Brian Parker, Matt LaCosse, and Dymonte Thomas have 2 credited seasons, so require tenders of $645K each. A little simple multiplication and addition tells me it will cost $4.365M to tender our 7 ERFAs. I subtract that from the $29.327M I calculated in the preceding paragraph, and I have $24.962M left. But that's NOT top 51 cap space. I'm now counting 60 players against the cap.

Accounting for money needed to sign draft picks:

It has now become rather obvious to me that any player with a cap hit of < $570K will fall outside of the top 51. I'll use Spotrac's projected cap hits for our draft picks that are > $570K. $3.409M (for 1-10) + $1.293M (for 2-41) + $.760 (for 3-71) + $.664M (for 4-119) + $.575M (for 5-139) + $.571M (for 5-147) = $7.272M for 6 players. I subtract that from the $24.962M I calculated in the preceding paragraph, and I have $17.69M left. But that's NOT top 51 cap space. I'm now counting 66 players against the cap.

Applying the top 51 rule/adding money back to our cap space:

I now get to add the 15 lowest cap hits I'm counting back to our top 51 cap space because 66 - 51 = 15. Spotrac shows us with 3 cap hits of $495K counting against the cap, but I already counted one of those fixing the Jimmy Williams mistake. The next lowest cap hit Spotrac shows is $565K. And then they show 10 cap hits of $570K, but I added 2 more of those when I tendered the ERFAs. $.495M + $.495M + $.565 + (12 X $.57M) = $8.395M. I add that to the $17.69M i calculated in the previous paragraph, and I have $26.085M in top 51 cap space.

Conclusion:

I've calculated an estimated top 51 cap space of $26.085M after cutting Keenum and Stewart, extending Kreiter, tendering 3 out of our other 4 RFAs, tendering all 7 of our ERFAs, and signing our draft picks. If we're miraculously able to trade Keenum instead of cutting him, add $7M to that number (his guaranteed $7M salary) and subtract $805K (the veteran's minimum salary for him in free agency I assumed). That would result in $32.28M in top 51 cap space. Maybe we can't trade Keenum, but we decide not to exercise our option on Sanders. What would that give us? $26.085M + $10.25M - $.57M = $35.765M. Basically, for every player we cut or trade who costs more than $570K, a $570K cap hit will take the place of his cap savings after dead money in the top 51. And for every more expensive free agent we sign, a $570K or higher cap hit will fall out of the top 51,