Queen's Gambit Inclined

THE Exchange Variation of the Queen's Gambit Declined tends to give White a long-lasting plus-over-equals.
It used to be very popular at the top level, but has fallen out of favour among most of the best grandmasters as it is thought a competent Black should be able to eke out a draw.
Different considerations apply in tournaments such as Benidorm U2000.
Spanton (1940) - Ramiro Rueda Maria (1791), round four
Queen's Gambit Declined, Exchange Variation
1.d4 d5 2.c4 e6 3.Nc3 Nf6 4.cxd5 exd5 5.Bg5 Nbd7
This sets a little trap (6.Nxd5?? Nxd5 7.Bxd8 Bb4+) that has ensnared more than 500 players in ChessBase's 2019 Mega database, but is generally a harmless transposition.
6.e3 Be7 7.Bd3 0-0 8.Nf3 Re8 9.Qc2 c6 10.0-0 h6 11.Bh4 Ne4 12.Bxe7 Qxe7

This tabiya occurs more than a thousand times in Mega '19

13.Nd2!?
There are more than 110 examples of this move in Mega '19, but more popular are 13.Rab1, 13.Bxe4, 13.b4!? and 13.Rae1.
The surprising move among those, at first glance, is 13.b4!?, but the point is that 13...Qxb4 can be met by 14.Nxe4 (14.Bxe4 is also possible) dxe4 15.Bxe4.
13...f5
Major alternatives are 13...Ndf6 and 13...Nxd2.
14.Rae1
Proceeding with a standard Minority Attack seemed unwise with a black knight ensconced on e4. At the same time, I did not want to capture on e4 and give Black even more kingside space.
14...Qg5?!
This has been played by a 2365, but it seems to leave Black on the back foot. The analysis engines Stockfish9 and Komodo9 prefer 14...Ndf6, 14...Nb6 or 14...Nxf2.
15.f3!?
Peter Dittmar (2352) - Wolfgang Gerstner (2365),  Bundesliga 1999, saw 15.f4 Qf6 16.Nf3 Nb6 17.Ne5, with a position where the engines much prefer White (although, according to the database, Black eventually won on time).
The engines like 15.Bxe4 fxe4 16.f3 exf3 17.Nxf3, with an initiative.
15...Nxc3 16.f4!
A similar idea to the Dittmar - Gerstner game, but without leaving a black piece on e4.
16...Qf6 17.bxc3 Nf8 18.Nf3 Be6 19.Ne5 Nd7

White has much-the-better minor pieces

There now follows a long series of moves in which I am able to combine pressure against b7 with pressure against Black's kingside.
20.Rb1 Rab8 21.Rb2 g6 22.Rfb1 Nxe5 23.fxe5 Qe7 24.g4
This could be regarded as double-edged in that both kings become somewhat exposed, but I felt White had more manoeuvring room in which to exploit the situation.
24...Rf8 25.gxf5 gxf5 26.Kh1 Kh7 27.Rg1 Rg8 28.Rbb1 Rxg1+ 29.Rxg1 Rg8 30.Rf1
Avoiding more simplification, while trying to reduce Black to passively defending the f5 pawn.
30...Rg5!?
The engines prefer this to the entirely passive 30...Rf8.
31.Qf2 Qg6
The engines suggest 31...Qe7, but reckon White keeps the upper hand.
32.h4 Rg3?
Afterwards - and during the game, actually - RM was angry with himself at not having played the move he intended, ie 32...Rg4, the point being that trying to win the exchange with 33.Be2?? loses to 33...Rg3, eg 34.Kh2 f4 etc.
But White instead plays the immediate 33.Kh2, eg 33...Qh5 34.Bxf5+ Qxf5 35.Qxf5+ Bxf5 36.Rxf5, with a much-superior rook-and-pawn ending, although the game goes on.
33.Kh2 Rg4 34.Be2 f4 35.exf4 1-0