Drillers cut six oil rigs in the week to March 29, bringing the total count down to 798, General Electric Co’s Baker Hughes energy services firm said in its closely followed report on Thursday.
Baker Hughes published its North American rig count report on Thursday, one day earlier than usual, due to the Good Friday holiday on March 30.
The U.S. rig count, an early indicator of future output, is much higher than a year ago when 662 rigs were active. Energy companies have been steadily increasing spending since mid-2016 when crude prices began recovering from a two-year crash.
U.S. crude futures traded around $65 a barrel this week, approaching the three-year high of $66.66 hit in late January, up sharply from the $50.85 average hit in 2017 and$43.47 in 2016.
Looking ahead, futures were trading around $63.50 for the balance of 2018 and $59 for calendar 2019.
In anticipation of higher prices, U.S. financial services firm Cowen & Co said 58 of the roughly 65 exploration and production (E&P) companies they track have already provided guidance indicating an 11 percent increase this year in planned capital spending.
Cowen said those E&Ps that have reported capital plans for 2018 expected to spend a total of $80.5 billion in 2018, up from an estimated $72.4 billion in 2017.
Analysts at Simmons & Co, energy specialists at U.S.investment bank Piper Jaffray, this week forecast the total oil and natural gas rig count would average 1,015 in 2018 and 1,128 in 2019, the same as last week.
So far this year, the total number of oil and natural gas rigs active in the United States has averaged 966, up sharply from an average of 876 rigs in 2017 and 509 in 2016, and not far from the total of 978 in 2015. Most rigs produce both oil and gas.
EIA projected this month that average annual U.S. production will rise to a record high 10.7 million barrels per day (bpd) in 2018 and 11.3 million bpd in 2019, up from 9.3 million bpd in 2017.
The current all-time U.S. annual output peak was in 1970 at 9.6 million bpd, according to federal energy data.