EXPERIENCE AND RESEARCH SHOW BEST DESIGNS FOR FOAM-DIVERTED ACIDIZING

R.D. Gdanski
Halliburton Services
Duncan, Okla.

Guidelines developed from field experience G and laboratory tests increase the probability of removing damage and stimulating production with foam-diverted acid treatments.

Foam for diverting acid has lately received more interest because of the success offshore in stimulation of relatively short intervals of gravel-packed sandstone reservoirs.1

Recent laboratory research has augmented the knowledge of foam diversion in limestone cores. 2-3 However, successful large-scale treatments require the consideration of a number of factors including:

Tubing size
Treatment volume
Stage size
Foam quality
Formation Composition
Porosity and permeability
Completion configuration
Pumping, rates
Interval length

FOAM TREATMENTS

During the past 10 years, foams for diverting stimulation fluids have had good successes in wells where solid diverters are either undesirable or too expensive.

The primary success has been in diverting hydrofluoric (HF) acid in gravel-pack screens and where formation thickness is less than 50 ft. The trapping of oil soluble resins and benzoic acid is of concern if poor quality products are used or if the materials are used outside recommended ranges.

A second success has been in HF treatments of very long intervals of sand and shale completed with gravel and slotted liners. These wells typically have 1,000-ft thick formations and are drilled vertically. Factors favoring foam are the cost of solid diverters and the lack of confidence of the effectiveness of solids over long intervals.

Foam slugs (Fig. 1) are common for short interval treatments while continuous diversion by foaming the entire stimulation treatment is prevalent for treating long intervals.

Most designs have not considered experimental data. Treatments typically have been pumped down either coiled tubing or small, such as 2-7/8-in, tubing.

Successes of foam diversion have encouraged researchers to investigate improving the process. Most research, however, has been in conformance control in enhanced oil recovery processes where gas injection follows foam injections But certain research directed toward stimulation has improved insight into the diversion process during stimulation. 1-3 5

Some research has modeled foam diversion on the same scale as stimulation treatments but without rock dissolution. Other work has focused on the impact of foam on dissolution properties in carbonates. 2-3

Foam has been used in very long horizontal carbonate intervals as well as long intervals of vertical sandstones. These primarily international wells have tubing configurations different from those in earlier treatments.

HF IN SANDSTONES

Design of foam diverted HF acid treatments of sandstone formations must first recognize the dissolution patterns. Specifically, one can view HF acidizing as an interstitial process. That is, the fluids dissolve clay and silicate materials as they flow around and through the porous media.

Wormholes are rarely generated, although, channeling undoubtedly occurs in some instances. In addition, the amount of rock dissolved is relatively small.

For example, sandstone stimulation with 100 gal/ft of 12% hydrochloric (HCI)-3% HF results in the dissolution of only 18 lb of rock/ft of perforated interval. Thus, 100 gal/ft of HF increases the void space around the well bore by less than I gal/ft of perforated interval.

Essentially all of the pumped stimulation fluid must occupy spaces within the unstimulated matrix. As a result, treating pressures usually remain high and the risk of fracturing can also be high.

Precautions are taken to avoid fracturing because formation fracturing during an HF treatment is considered to be catastrophic for successful stimulation.

Fracturing occurs if the permeability is unable to accept the foam at adequate rates. Indeed, foamed fluids are excellent for proppant fracturing because of excellent fluid-loss control.

This leads to the first important guideline concerning the use of foam diversion. Foam is generally not used in sandstone formations having permeabilities less than about 50 md. The permeability range of 50-130 md is one that requires careful design and caution if the formation is treated with either too much foam or too high-quality foam.

Sandstone formations with permeabilities above 150 md seem to have little risk and are the best candidates for foam diversion of HF acid treatments.

INTERVAL LENGTH

The interval length treated determines the acid volume. Traditionally, volumes are 50 gal/ft of HCI preflush, 150 gal/ft of an HF acid stage, and 50 gal/ft overflush,

For even, 100-ft interval, this equals 5,000 gal of preflush, 15,000 gal of HF acid, and 5,000 gal of overflush.

The first decision is whether the entire perforated or open hole interval will take fluid and/or produce fluid. In some cases, 50% of the interval may be impermeable shale and can be ignored. In other cases, the entire interval is potentially productive.

The total treatment volume should be based on the length expected to take stimulation fluid.

The length is also important for determining the surface pumping rate. A general starting point for pumping rates that should avoid fracturing is about 0.05 bbl/min/ft or 5 bbl/min/100 ft. But if coiled tubing is used to treat a 100-ft internal, this rate would be too high because of excessively high friction pressures. Even with 3-1/2-in. tubing, the rate could be too high because of formation damage. The maximum rate would likely be about I bbl/min.

A 500-ft interval and 3-1/2-in tubing indicates a rate of about 25 bbl/min. But this rate might be impractical because of logistics and formation damage.

A more practical pump rate would be about 10 bbl/min.

TUBING SIZE

Once the pump rate is determined, two times are calculated: transit time from the surface to the first perforations and total treatment time.

Transit time is important for judging whether the treatment can be made and depends on the tubing size.

For example, pumping at 1 bbl/min through coiled tubing with 10 bbl capacity results in a transit time of 10 min. If the foam stage provides too much resistance in the formation, the quality or size can be reduced.

The type of formation damage can be indicated by knowing if the break in treating pressure occurs when HCI or HF is on the formation.

In addition if the first acid stage quickly goes on a vacuum, increasing the volume of foam may improve diversion. In large diameter tubing this flexibility is lost because of long transit times and large fluid volumes.

The total treatment time is important for determining how long the initial foam stages must provide resistance. Treatments expected to last for over 5 hr should have large enough foam volumes in subsequent stages to replace previous stages.

Tubing size also determines the volume for individual stages. With a 10-bbl capacity coiled tubing, stage volumes are rarely restricted by tubing capacity. With large diameter tubing however, serious consideration should be given to the effects of mixing during transit.

For example, the capacity of 12,000 ft of 4-in. tubing is about 140 bbl. For a 100-ft interval, one might use five diverted stages. Breaking the preflush HCI into 1,000 gal increments would mean that the first stage occupied only 17% of the tubing.

With a 3 bbl/min pump rate, the transit time is about 45 min. Therefore, it seems reasonable to expect significant mixing between the separate stages. In fact, a mixing zone of about 10% of the tubing capacity is reasonable because of the long transit time, density difference with the following HF acid, and thermally induced mixing effects in large-diameter tubing.

For a stage, a minimum volume of at least 30% and preferably 50% of the tubing string seems more reasonable. In the same example, the HCI preflush of 120 bbl could be split into two stages consisting of 30 and 70 bbl. The first stage would occupy 36% and the second stage 50% of the tubing volume.

A three-stage preflush would result in either three equal stages of 40 bbl (29% of the tubing volume) or a significantly smaller initial stage. Thus, the conclusion is that two is the maximum number of sequences without causing serious mixing.

FOAM SLUGS

After establishing the number of diverted sequences, the next step is to determine the foam diverter stage. The two sequences will be separated by a single diverter stage.

If the ratio of the sequence volumes is applied to the interval length, the first sequence will treat about 40 ft. Although a little long, this is still a reasonable interval length over which one might expect the first stimulation sequence to enter the formation.

For determining the foam slug volume, a general guideline from field experience is I bbl of foam for each foot of diverted interval. Thus in this case, the foam slug should be about 40 bbl.

One must, however, also consider the volume occupied by the foam slug in the tubing. For 12,000 ft of 4-in. tubing, 40 bbl of foam occupies 29% of the tubing. Therefore, although the stage is probably large enough from our general guideline, the large difference in density between the foam slug and the stimulation fluids will likely be a major problem.

With a transit time of 45 min at 3 bbl/min, the foam slug may mix significant because of the density gradients. For small diverting foam slugs in large-diameter tubing, maximum transit times probably should be about 20-30 min.

The options are:

Accept the mixing
Pump the treatment through coiled tubing
Pump at a faster rate
Pump a larger foam slug
Use the continuous diverting technique.

Experience with the formation can indicate the best option.

CONTINUOUS DIVERSION

As an example of a continuous diversion, assume a well with a 500-ft, 100-300 md sandstone and 12,000 ft of 5-in. tubing, 225 bbl capacity

Based on the interval length, treatment volume is:

Acid preflush: 25,000 gal (600 bbl)
HF acid: 75,000 gal (1,800 bbl)
Overflush: 25,000 gal (600 bbl).

With foam slugs, the treatment might be broken into 10 stages covering 50 ft each. In this case, the 50-bbl foam slugs would only occupy about 20% of the tubing. Therefore, because of the large tubing volume, the foam would have little chance of reaching the perforations. In such cases, continuous diversion with increasing foam quality is advantages.

Because field experience has shown that nitrogen and HF of 40 to 50 quality has little diverting effect, the first sequence of preflush, HF, and overflush should be at least be 30-quality foam. Also because 70-quality foam can produce too much resistance, the last sequence should not be greater than 70 quality.

Next, if one assumes that 5% increments of quality will increase resistance to matrix flow and thus diversion, the 5 sequences should have 50, 55, 60, 65, and 70 quality foam.

It is reasonable to expect that initially higher permeability intervals are treated. These sections are also likely to require less stimulation or damage removal.

Thus, the first sequence will be the smallest with 10% of the volume. If each subsequent sequence is increased by 5% of the total volume, the five sequences will have 10, 15, 20, 25, and 30% of the treatment volume.

Therefore, the first sequence will be as follows: *Acid preflush: 60 bbl 0 HF stage: 180 bbl *Overflush: 60 bbl.

The 60-bbl of acid when foamed to 50 quality will result in 120 bbl. This volume occupies 53% of the tubing and is well within the guideline for a minimum stage volume.

The second acid preflush of 90 bbl when foamed to 55 quality increases to 200 bbl of foam. Because this acid preflush follows the first overflush, the higher quality foam's lower density and higher bulk viscosity prevents the acid from mixing as readily.

Thus, the continuous diversion with increasing quality has the benefit of reducing the detrimental mixing effects. In fact, the entire treating volume could be halved and still be within the guidelines for stage volumes.

The continuous treatment can be further modified based on well response. If a sequence pressure becomes too high, the increased volume because of the foam allows the flexibility during the next sequence to have the same or a slightly reduced quality, if necessary.

Of course, if the treating pressure increases above the designed pressure, then compression of the gas can reduce downhole foam quality and ease formation entry. In other words, various treating parameters can provide an almost self-adjusting foam quality.

HCI IN CARBONATES

Foam diversion in HCI treatments of limestone formations must first recognize the dissolution patterns. Specifically, carbonate acidizing can be viewed as a process that generates wormholes and interconnects natural fractures and vugs. That is, the fluids dissolve the matrix and leave open voids.

The void space is equal to the dissolved rock volume plus the porosity. A quantity of 1,000 gal of 15% HCI will dissolve 10.9 cu ft of limestone. If the porosity is 10% then the total void space will be about 12 cu ft.

Because 1,000 gal occupies 134 cu ft, this indicates that over 90% of the pumped fluid must leak off into the unacidized matrix. This leakoff process helps generate uncountable mini and micro-wormholes.

Research has shown hairlike wormhole patterns branching off larger complex wormhole networks. As a result, diversion in limestone matrix acidizing can be quite difficult to achieve.

The most common methods uses either ball sealers that plug, perforations or a "block" that is a suspension of solids in a gelled carrier fluid. There are times, however, when it is impractical or logistically difficult to use ball sealers or other solids. One such case is in acidizing a very long horizontal well.

One feature of matrix acidizing carbonates is that occasional fracturing after a diversion stage is quite acceptable. While it is generally not intended to treat the formation entirely above fracturing pressure, pressure increases from a diversion stage can fracture the damaged sections of the matriy thereby opening up flow passages.

Because leakoff will still be massive, the treatment will quickly return to matrix acidizing and the treating pressure will fall below fracturing pressure. This feature minimizes the risks of excessive foam diverting stages and alloys enhanced design flexibility,.

In addition, recent research indicates that foamed acid tends to produce less branched and longer, narrower wormholes than nonfoamed acid. This effect allows for deeper stimulation from the well bore with smaller volumes of acid.

A positive factor for foam in acidizing chalks, which have minimal compressive strength to start with, is that the matrix surrounding the well bore receives deeper stimulation with less rock removal.

PERMEABILITY

Formation permeability is the first consideration in designing foam diverted acidizing of carbonates.

Carbonates with low permeabilities, less than 1 md, and low porosities, less than 5%, will likely need small foam stages. In these cases, the acid needs sufficient opportunity to penetrate the carbonate and develop the wormhole patterns. Also, the foam will easily divert the acid.

Carbonates with medium permeabilities or which have moderate natural fracturing can be diverted with either large foam stages or with continuous foam. Such carbonates respond quickly to acid and thus require more diverting resistance as the foam penetrates the natural fractures and matrix permeability.

Carbonates with high permeabilities, above 50 md, or high porosities, above 25%, generally require continuous foam diversion. Foam invasion of the matrix can be significant and thus diversion will require more resistance to flow. Indeed, while low-permeabibty carbonates can be diverted with foamed acid, high-permeability carbonates will require foamed brine or even foamed gelled brine to achieve a lasting diversion.

Finally, some formations possess a thief zone of "hyper" permeability. These zones require a temporary block such as a delayed crosslinked, high-viscosity gelled water as the initial treatment stage.

These temporary blocks can also contain internal breakers to control the break of the gell as well as contain large solids such as oil soluble resins, graded salt, or graded calcium carbonate to establish and strengthen the block.

Once the temporary block is in place, then continuous foam diversion should be used to acidize the rest of the formation.

INTERVAL LENGTH

Interval length is again used to start treatment design. Typical treatments for matrix acidizing a carbonate range from about 25 gal/ft for a small perforation wash to remove or dislodge damage to about 230 gal/ft for deeper penetration and stimulation.

Economics and logistics can also play an important role in determining the acid volume. If 50 gal/ft are used for a long interval, than 5,000 gal are needed for every 100 ft of treated interval.

A reasonable treating rate is also based on interval length. Because carbonates respond much more quickly to acid and occasional fracturing is acceptable, a reasonable starting point for a surface pumping rate is 0.1 bbl/min/ft of treated interval.

At first, all of the interval may not take fluid. For example, in a 500 ft interval if only 100 ft at a time takes fluid then the surface treating rate would be set at about 10 bbl/min.

Low-permeability carbonates might require lower rates while high-permeability might tolerate higher rates.

One would like to pump as fast as reasonably possible without frequent fracturing. Too much fracturing will prevent the desired diversion.

TUBING AND CASING

In carbonates, the stage size is easier to determine. The most obvious reason is the simplicity of the fluid sequences. Unlike HF acidizing that has three fluid stages per sequence, carbonate acidizing has one acid and perhaps a non-acid diverting fluid.

Therefore, to achieve the minimum requirement of 3050% of tubing volume is relatively straightforward.

The transit time and total treatment time are still calculated for judging effectiveness and obtaining the duration of individual diverting stages.

Because many horizontal wells are completed in carbonates, the dimensions of the casing or liner in the horizontal section need to be known. It was previously assumed that the foam slugs after exiting the tubing would enter the formation. This assumption is adequate for vertical wells and intervals shorter than a few 100 ft.

In horizontal wells, however, gravity induced overriding can be a serious problem. Therefore, the diverting stage needs to be large enough to be effectively pumped for long distances in large-diameter pipe.

For example, the capacity of 100 ft of 7-in. casing is about 4 bbl. While this volume of foam is reasonable if pumped through coiled tubing, it would not be reasonable to expect it to travel in 1,000 ft of horizontal hole without segregating and dispersing.

Once again, the guideline for the minimum volume is 30-50% of traveled capacity. That is, a 4-bbl stage of foam could be expected to travel about 200-350 ft and still remain fairly intact. However, as more stages are diverted farther in the horizontal section, the size must increase to maintain foam diverter integrity.

For example, the final stages for a 2,000-ft horizontal well with a 7-in. liner, capacity 80 bbl, should be a minimim of 25-40 bbl of foam diverter.

FOAM SLUGS

The size of a diverting foam slug is determined based on the acid stage pumped just before it and the recommended minimum slug size for the horizontal distance traveled.

First, the void space created by the preceding acid stage is calculated. With 5,000 gal (120 bbl) of 15% HCI and a porosity of 20%, Table 1 shows that the void space is about 12 bbl.

Therefore, a minimum of 12 bbl of foam (10% of the acid volume) is required to fill the void. Additional foam volume is needed because of the possibility that:

The leading 25-50% of the foam enters the matrix beyond the created wormholes
The first portion of foam entering the matrix is broken by residual oil, surfactant adsorption, or other effects.

In a low-permeability carbonate that is expected to produce dry gas, 25% excess foam would probably be sufficient. However, in a high-permeability or high-porosity carbonate expected to produce oil, 100% excess would be quite reasonable.

The foam slug, therefore, would range from about 12 to 20% of the acid volume. Because 28% HCI has twice the rock dissolving power of 15% HCI, this range would double for 28% HCI.

If the preceding acid stage was 5,000 gal of 15% HCI in a formation with a porosity of 20%, one would probably use 100% excess or a total of 25 bbl of foam (21% of the acid volume).

This volume should be increased if it is below the guidelines for the minimum volume. If the diverter stage is increased by more than 50%, then one should also consider increasing the preceding acid stage by a proportional amount.

Once the foam diverter is in place, it Will continue to invade the matrix and the foam will gradually break. Thus, the amount that each individual foam diversion stage breaks or is lost by penetration into the matrix should be account for.

Very little information is available on this subject. The lost foam, however, will likely be replaced first as the successive stages of foam diverter are pumped past the interval.

A reasonable estimate of this loss would be about 1025% of the previous foam volume occupied in the created voids. The successive foam diverter stages should be increased to account for this loss.

For example, if the second stage of acid was also 5,000 gal (120 bbl) of 15% HCI, one would again design 25 bbl of foam for the diverter stage. Because the first acid stage created a void of about 12 bbl, an additional 1-3 bbl would bring the total to 2628 bbl.

Laboratory work has indicated that foamed acid can give excellent resistance and diversion. 2 Logistically this is beneficial because only increases in nitrogen rate are required for high-quality foam. Switching to a different fluid is not necessary.

Because 70 to 80-quality foams would be recommended for the foam diverting stage, 50 bbl of foamed acid would require only 15 and 10 bbl of acid, respectively, to create the foam diverter. To help prevent gas dispersion from the foam diverter stage, the remaining acid could be commingled with nitrogen to about 40 quality .

This approach works well in low to medium-permeability carbonates. In high permeability or high-porosity carbonates, one should primarily consider continuous foam diversion.

Logistics, however, can dictate that foam slugs must be used. In such cases the foamed diverter stage should be a non-acid such as freshwater, a light brine, or sea water. Non-acid diverter stages would also be recommended in high porosity chalks as well, regardless of the formation permeability.

Laboratory studies using chalks have shown that foamed acid will penetrate rapidly through a high-porosity chalk resulting in minimal diversion. Tests conducted with foamed brine in the same chalks have shown large buildups in flow resistance that were largely retained even after nonfoamed acid followed the foamed brine.

The ineffectiveness of foamed acid in providing diversion in high-porosity chalks was presumably because of the small grain volume present in a given bulk volume of chalk as compared to an equivalent bulk volume of a low-porosity limestone.

In low porosity, medium-permeability limestones, foamed acid generally gives adequate diversion. Gelling agents can add stability in the foam stage.

Because gelled foam will be displacing the acid from the wormhole network, an exceptionally stable gelling agent in acid is not required. Natural polymers such as xanthan, derivatized guar, and derivatized cellulose are acceptable.

Gelled foam stages would be considered in high volume, low-rate treatments such as those through coiled tubing in horizontal wells. Such treatments can last for many hours and require the additional foam stability provided by the gelled base fluid.

CONTINUOUS DIVERSION

High-permeabilities carbonates should primarily be treated with continuous foamed acid. As in HF acidizing, the quality can start low, 50 quality, and be gradually increased. In carbonates, however, the foam quality can be increased to 80 quality, if necessary.

The goal of continuous foam diversion is to provide sufficient resistance to matrix flow so that all sections of the interval are exposed to acid. If the permeability is sufficiently high, one can start with 60 quality to provide quick diversion. One of the primary advantages of continuous foam diversion is that calculation and control of stage size is no longer necessary.

Stages of foamed non-acid fluids can also provide enhanced diverting effectiveness in continuous diversion treatments. In such cases, one must balance the benefits for the stimulation with the logistics of transporting additional fluid, etc.

This approach would be used with constant-quality foamed acid at about 60-70 quality with stages of foamed diverter designed using the foam slug technique. This technique is the method of choice for diversion in high-porosity chalks.

REFERENCES

Burman, J.W., and Hall, B.E., "Foam as a Diverting Technique for Matrix Sandstone Stimulation," Paper No. SPE 15575, SPE Annual Technical Conference and Exhibition, New, Orleans, Oct. 5-8, 1986.
Thompson, K. and Gdanski, R. D., "Laboratory Study Provides Guidelines for Diverting Acid With Foam," Paper No. SPE 23436, Eastern Regional Meeting, Lexington, Ky., Oct. 7-25, 1991.
Bernadiner, M.G., Thompson, K.E., and Fogler, H.S., "The Effect of Foams Used During Carbonate Acidizing," Paper No. SPE 21035, International Symposium on Oilfield Chemistry, Anaheim, Calif., Feb. 20-22, 1991.
Prud'homme, R.K., and Khan, S ed., Foams, Theory, Measurements and Applications, Marcel Dekker, Inc.
Zhou, Z.H., and Rossen, W.R., "Applying Fractional-Flow Theory to Foams for Diversion in Matrix Acidization," Paper No. SPE 24660, 67th Annual Technical Conference and Exhibition, Washington, DC, Oct. 4-7, 1992.
Hoefner, M.L., and Fogler, H.S., "Fore Evolution and Channel Formation During Flow and Reaction in Porous Media," AlChE Journal, January 1988, pp. 45-54.