I've looked at [B]18 a4 bxa4[/B] a bit more.

[QUOTE=cheesehead;364663]
[B]18 a4[/B]

18 ... bxa4
19 Rxa4 Ba6 20 Nd2 wins the c-pawn.[/QUOTE] 20 ... Bb5 says, "Not so fast", but after 21 Ra2 (perhaps 21 Bxc6+ Nxc6 could be interpolated before 22 Ra2 -- does this have some advantage for us?) we can pile pressure on c4 (N-e4-d6, Bd2 and Ne3, then Rfa1 doubles our rooks) and we might hope to make Black regret having pushed his Q-side pawns so far. Of course, this does give Black plenty of time to castle or otherwise bring his R/h8 into play.

[QUOTE=cheesehead;365123]we can pile pressure on c4 (N-e4-d6, Bd2 and Ne3[/quote]One tempo faster, after Ra2 protects the b-pawn, is N-f2-e4-d6 with no B move required. (The N/d2 already attacks c4, so doesn't need to move for that purpose.)

That doesn't mean we don't have to carefully consider our Q-side pawn situation all through future deliberations. But Nf2 makes Rfb1 a reasonable one-tempo alternative to consider.

If 18 ... bxa4 then 19. Nd2 Ba6 20. Rxa4 Bb5 21. Ra2 followed by Ne4

[QUOTE=WMHalsdorf;365205]If 18 ... bxa4 then 19. Nd2 Ba6 20. Rxa4 Bb5 21. Ra2 followed by Ne4[/QUOTE]Yummy.

Could Black at some point decide to sac a piece for our two advanced pawns?

From the position after 21 Ra2 above:

21 ... Bxe5 22 dxe5 Nxe5 23 Bf4 N7c6 24 Bxc6+ Bxc6 25 Bxe5
or 23 ... f6 24 Re1 N7c6 25 Bxc6+ Bxc6 26 Bxe5 fxe5 27 Rxe5

21 ... Nxe5 22 dxe5 Bxe5 23 Rxa5 B/e5 retreats somewhere (not 23 ... Nc6 24 Bxc6+ Bxc6 25 Rxe5) because of threats like 24 Re1 and 25 Bxh6 to win a pawn or 24 Nxc4 Bxc4 25 Rxe5 Bxf1 26 Kxf1 and White has two bishops for the rook (and 2 connected passers vs. only 1 passer for B).

(Now I'll go back to analyze possible Black sacs of piece for our d- and e-pawns earlier.)

The sac 21 ... Bxe5 22 dxe5 Nxe5 would be netter answered with Bd4 rendering f6 unplayable. We could also skip move 20. Rxa4 and go straight to 20. Ne4

[QUOTE=WMHalsdorf;365222]The sac 21 ... Bxe5 22 dxe5 Nxe5 would be netter answered with Bd4 rendering f6 unplayable.[/quote]Thanks!

Arggh! I did look at 23 Bd4, and initially wrote down my analysis as 23 Bf4(or Bd4) N7c6 24 Bxc6+ Bxc6 25 Bxe5, but then noticed the "or 23 ... f6" line after Bf4, so I erased the "(or Bd4)" ... intending to write out the 23 Bd4 analysis separately from Bf4 ... but I forgot!

[quote]We could also skip move 20. Rxa4 and go straight to 20. Ne4[/QUOTE]... and of course I forgot to write down that line, too after I'd looked at it!

18 ... bxa4 19 Nd2 Ba6 20 Ne4
Now if 20 ... O-O 21 Nc5 (threatens Nd7 rook-fork) Bc8 22 Rxa4
or 21 ... Ra8 22 Nxa6 Rxa6 23 Rxa4 (threatening 24 Rxc4)

I have to leave now, before analyzing 23 ... Bxe5 24 Bxh6