Halo Effect Revisited -- May 5, 2015

Update

May 6, 2015: after writing all of that below, there's always the possibility of mis-allocation if production was commingled; I'm not going to follow-up on this for now; just something to be aware of; [a reader reminded me of that some time ago in a different post: large production number could be due to 'mis-allocation' during early commingling; random update, March 20, 2015]. 

Original Post

This is why I love to blog about the Bakken.

I don't think I've blogged about this one before; if I have, I've forgotten.

This is really, really cool.

First, look at this production profile, only the last eleven months or so:

Pool	Date	Days	BBLS Oil	Runs	BBLS Water	MCF Prod	MCF Sold	Vent/Flare
BAKKEN	3-2015	19	5150	5252	667	11562	11328	197
BAKKEN	2-2015	28	8427	8442	1097	19345	19025	264
BAKKEN	1-2015	31	12082	12158	1973	25910	20753	5095
BAKKEN	12-2014	31	15002	14875	2250	29428	17606	11760
BAKKEN	11-2014	29	12289	12438	2604	20878	2377	18446
BAKKEN	10-2014	17	7005	6713	1991	11756	1782	9942
BAKKEN	9-2014	0	0	0	0	0	0	0
BAKKEN	8-2014	18	3980	3992	401	7959	7745	179
BAKKEN	7-2014	31	7128	7222	373	14570	13727	781
BAKKEN	6-2014	30	8107	8067	318	15769	12953	2756
BAKKEN	5-2014	31	8503	8480	361	14081	12934	1085

Notice anything unusual about it? Yup, in August, 2014, it was taken off-line and was put back on-line in October, 2014. When it went off-line it was producing 7K bbls/month and production was decreasing.  When it went back on-line in October, it produced 12K and then 15K the following month. Four things can explain the increase (and a pump is not one of them):

• the well was re-worked
• the well was re-fracked
• the halo effect of a neighboring well being fracked
• "build-up" which occurs when a well is taken off-line

The production profile above was from this well:

• 20589, 4,815, Whiting, Tarpon Federal 21-4H, Twin Valley, middle Bakken, t10/11; cum 518K 3/15; 

This well was fracked back 10/10/11 (October, 2011). It was not re-fracked in September, 2014, to explain the jump in production.

There is nothing in the file report to suggest why the increase in production in October 2014. It was not a work-over rig and it was not re-fracked; I can say that with almost 100% certainty; I'll explain later.

One possibility is the "halo effect" of a neighboring well being fracked. So, let's check the nearest horizontal well (laterally, not vertically). That would be:

• 27520, 6,002, Whiting, Flatland Federal 11-4TFH, t10/14; cum 311K 3/15;

Pool	Date	Days	BBLS Oil	Runs	BBLS Water	MCF Prod	MCF Sold	Vent/Flare
BAKKEN	3-2015	31	46708	46286	4808	181103	180141	900
BAKKEN	2-2015	28	46881	47420	4187	146103	143797	2250
BAKKEN	1-2015	31	52430	52198	4093	141901	128516	13323
BAKKEN	12-2014	31	52103	51904	3664	123971	115165	8744
BAKKEN	11-2014	30	50823	51268	4155	112069	42428	69581
BAKKEN	10-2014	31	61727	60615	12213	117291	11381	105860
BAKKEN	9-2014	27	0	0	0	44	0	0

And, yes, indeed, it was fracked in the September, 2014, time frame which explains why # 20589 above was taken off-line.

The only problem: #27520 is a Three Forks well. It's hard to believe that fracking a Three Forks well would affect a middle Bakken well, at least according to some / many.

So, we move to the middle Bakken well that was fracked at the same time:

• 27521, 5,002, Whiting, Flatland Federal 11-4HR, t10/14; cum 259K 3/15;

Pool	Date	Days	BBLS Oil	Runs	BBLS Water	MCF Prod	MCF Sold	Vent/Flare
BAKKEN	3-2015	31	26998	26811	3157	100230	99670	498
BAKKEN	2-2015	28	31865	32428	3174	92376	90898	1422
BAKKEN	1-2015	31	45023	44921	3898	104257	94408	9787
BAKKEN	12-2014	31	50609	50398	4036	111706	103765	7879
BAKKEN	11-2014	30	47947	48334	4104	98162	37160	60942
BAKKEN	10-2014	30	56457	55440	15512	129670	12583	117038
BAKKEN	9-2014	27	0	0	0	44	0	0

By the way, what happened to another older well in this same area when these two new wells were fracked? Take a look at #22361; it was taken off-line at the same time as #20589, when #27520 and #27521 were fracked. When it came back on-line, production was significantly increased:

• 22361, 4,971, Whiting, Tarpon Federal 21-4-3H, t12/12; cum 363K 3/15:

Pool	Date	Days	BBLS Oil	Runs	BBLS Water	MCF Prod	MCF Sold	Vent/Flare
BAKKEN	3-2015	31	7102	7089	862	17251	16895	294
BAKKEN	2-2015	28	6708	6684	848	16108	15832	220
BAKKEN	1-2015	31	8045	8094	1176	18163	14533	3568
BAKKEN	12-2014	31	9929	9900	1611	20821	12446	8313
BAKKEN	11-2014	28	11105	11506	1385	20222	2302	17864
BAKKEN	10-2014	22	11284	10777	666	19804	3003	16759
BAKKEN	9-2014	0	0	0	0	0	0	0
BAKKEN	8-2014	19	4688	4700	630	11281	10990	255
BAKKEN	7-2014	31	8257	8339	313	20234	19086	1086
BAKKEN	6-2014	30	8204	8152	326	20321	16707	3554
BAKKEN	5-2014	31	8013	8022	374	22733	20917	1754
BAKKEN	4-2014	30	8971	8962	433	25371	24510	80

Pretty nice bump up in production, huh?

I think I read somewhere that when a well is taken off-line, its production jumps when it comes back on-line simply because of the "build-up" that occurs. That could explain everything.

It's hard to believe that the "build-up" could result in production by that amount - it seems to be a pretty healthy increase.

But then note this: the increase in production, for whatever reason, is very short-lived -- only two to three months.

In addition to the "explosion" what else is required for a successful frack? Proppant. The propannt is there to help maintain the pores or the channels. A neighboriing frack does not put additional proppant into the neighboring well.

Whether the jump in production is due directly to a neighboring frack or is due simply to a "build-up" while the index well is off-line, the fact is that we see a jump in prodution.

Now, think about that.

During the slump in the price of crude oil (since October, 2014), Bakken operators are circling the wagons, putting their wells in the best spots from Williston to Watford City to the Parshall area. In other words, the wells that have been drilled in the past few months and likely to be drilled in the next twelve months are concentrated where older Bakken wells already exist.

In addition to "circling the wagons" (concentrating the wells to a more limited area), the operators are delaying completion (fracking) of these wells. There may be upwards of 1,000 wells waiting to be fracked (well above the expected 250 wells).

Obviously these 1,000 wells won't be fracked all at once, but we're going to see a lot of new wells fracked with newer techniques (including much more pressure and proppant) in an area already congested with producing wells sometime in the next twelve months.

Even if the production increase per well lasts only two to three months, there could be a significant increase in production overall in older Bakken wells simply because:

• there are so many wells already there, and
• these wells are in some great areas

Part of the reason some of these areas are so good is because of natural fractures; if so, the "halo effect" should be even more evident.

To complete the narrative, earlier I said: the increase in production was not a work-over rig and it was not re-fracked; I can say that with almost 100% certainty.

They don't do work-overs on wells near a well that is being fracked; in fact, they take those wells off-line completely (as they did above) for safety reasons. Also, there has been minimal interest in re-fracking in the past few months with the slump in prices; these wells were not coincidentally re-fracked. 

My narrative may be completely off base; that's fine. But the numbers don't lie.

Total vertical depth of three of the wells:

• 27520: 10,678 feet (Three Forks)
• 20589: 10,579 feet (Middle Bakken)
• 22361: 10,560 feet (Middle Bakken)

A graphic of the wells under discussion:

The distance between #27521 and #22361 is 0.68 miles, about 3,600 feet. That distance certainly speaks against a "halo effect" but if there is already a lot of natural fracturing in this area ...

... don't get hung up on whether this is due to the "halo effect" of fracking. The bigger point is the potential for a significant jump in production from older wells as the backlog of unfracked wells start getting fracked, regardless of the reason.

Note: I think I have the laterals identified correctly, but I do make mistakes more often than not. If I find I've made a mistake, I will update the graphic.