Friction Reduction in Tight Gas Stimulation
Mike Hurd – Technical Applications Manager, Kemira Oil & Mining
Global natural gas use is increasing as an efficient environmental alternative to oil and other liquid fuels. Production of tight shale gas formations is also growing to meet that demand and the related stimulation technology and chemistry required is developing quickly too. Heavy stimulation of these wells is needed and a significant part of that process is the chemistry used specifically for friction reduction. The efficiency of a hydraulic fracturing (frac) or acid job is directly related to the performance of the friction reducer (FR) on location and a surprising number of elements can affect that performance. The following article will review the major chemically related contributors to FR performance principally in ‘slick water’ type frac work where limited sand transport capacity is required. It concerns performance at the wellhead both during the treatment and where lasting effects can be seen.
Viscosity and Reynolds
Viscosity is actually the driving force behind friction reduction and molecular weight is the principle factor in the generation of viscosity. While there are some other factors involved, in general, whatever affects molecular weight effects friction reduction in the same way. Reynolds Number is one measure of this phenomenon since turbulent flow generates the highest friction in a flowing system. In a simplified Reynolds Number formula, R = DρV/μ, viscosity in the denominator has a great effect on the resulting degree of turbulence and therefore on the friction seen in the pipe. That said, calculations at the desk are one thing, but there are so many great, new products coming out that could become the components for a new frac fluid that some lab testing should be done. Viscosity data is good as interactions will show up there. Flow loop data through a fixed system though is the best, as it measures more closely the actual field events. There is debate on which setup is better, but there is no debate that loop data is better than anything else.
Polymers
The two chemistries normally used as friction reducers are guar (usually hydroxypropyl guar – HPG) and polyacrylamide (PAM) polymers. Variations and grades exist for both of these products that will modify performance somewhat, but improvement usually comes with a price. Since typical dosages range between 250-750 ppm (or 0.25-0.75 gpt) with thousands of cubic meters of fluid being pumped, more expensive options add cost to the treatments very quickly. Each of these polymers also has a couple physical forms that are often dependent on the freight from the manufacturing point or warehouse to the treatment location. Lower solids liquid or emulsion product is much easier to handle and use on location, especially if the footprint of the location allows for the inventory. Dry products are less costly in freight to location and may be the only option for platform inventory, but command additional equipment to put them in solution for use in the frac fluids.
Hot, Cold, and Salty
Similar products are already used in other areas of the oilfield and some of these characteristics may already be familiar there. Water salinity and well temperature are commonly cited as the biggest culprit to performance from typical polymers used elsewhere in the oilfields. There are two big temperature concerns here with hot AND cold water. Cold make-up or surface water causes a delay in putting the polymers in solution. ‘Time to solution’ can be a critical factor since water at a temperature <5°C can require twice the time to put emulsions, dispersions, slurries, or powders into a pumpable solution, than they would at 25°. Liquids can invert or dilute relatively quickly, but cold water delays that reaction and hydration, often by several minutes for each step. Yet both of these must occur for the polymers to reach full viscosity in very cold water. If it only takes 20 minutes for the pumped fluid to reach bottom then every minute delayed in the hole reduces the benefit of the polymer as a friction reducer. Since warmer water or a method to produce it isn’t always available, emulsion or liquid polymers specifically designed for cold water inversion can offset any additional product cost by reducing the need for additional equipment on location to compensate for the loss in performance. Such as higher horsepower, or to work around the delay for hydration in the form of additional blenders and storage.
Dry polymers have the same hydration problem as emulsions in cold weather since hydration is also delayed here, but the consequences can be more dire. The timeline for hydration is considerably longer for dries than for liquids and emulsions, especially in salty waters. Un-hydrated particles of a dry polymer can lodge in the wellbore and plug off injection during the treatment, remaining even after flowback as formation damage. Various grades of polymer can have different particle sizes and that can also impact the hydration timeline. The larger the particle, the longer it takes to go into solution. Cheaper grades of polymers may also have higher insolubles in them which never go into solution regardless of temperature or salinity, but act the same as an un-hydrated particle of polymer in terms of formation damage. Finer grinds of higher quality dry products and better hydration techniques on location may be additional costs, but they are essential to treatment performance should the decision be made to use dry products.
We need to go back and pick up the discussion of hot conditions, too. High bottom-hole temperatures can also cause significant problems and need to be addressed. There are realistic limits on both guar and PAM’s. Guar can handle up to 100-125°C in fresh water while PAM’s can perform well in the same range. The additional conditions of the fluids also matter. Salinity, hardness, and pH all become more critical as temperature goes up. Here is where some of the modified polymers work well and the added expense may well be worth the cost. AMP’s copolymers of PAM’s are resistant to both higher temperature and salinity effects while additives and crosslinkers in guar can also boost their performance under these conditions. There is continued debate on the significance of bottom-hole temperatures and most models suggest that the high rate of surface water injection will serve to cool the reservoir several degrees and protect the fluids being pumped to some extent. If the bottom-hole temperature (BHT) is within 25 degrees of the perceived limits of the polymer the injection rate, particularly with colder surface water in winter, will save you the cost of higher temperature products. If the BHT is beyond 50 degrees over the product limits you’ll definitely need to have a high temperature package and plan.
As suggested above, salinity, defined by Total Dissolved Solids (TDS), plays an important role in the development of the polymer’s viscosity. Different elements that make up the term ‘salinity’ have a different impact on the polymers themselves. Calcium limits the potential of both types of polymer to fully hydrate and build viscosity.
In careful lab observations of fresh water systems the effect can be seen on viscosity in as little as 50 ppm Ca. But performance effects in the field with all the other additives in the system aren’t typically seen until 100-400 ppm Ca is reached. Soda ash can complex the Ca and reduce the effect if it doesn’t interfere with other additives. In higher TDS systems (above seawater) chlorides tend to overtake the calcium as the problem forcing the basic PAM molecule to collapse on itself rather than hydrate fully. Some PAM polymers are being developed that can withstand a higher degree of salt, but guar tends to be less effected by monovalent salinity overall.
Biology
An additional concern that is growing in importance is biological activity and its relationship to these polymers in friction reduction. On the front end of these jobs there is a concern about ‘bugs’ in the surface water. Guar is particularly susceptible to a poorly designed biocide program with many species considering it a nutrient. Without a biocide program in place guar can lose viscosity within a matter of minutes depending on the bug population in the surface waters. PAM polymers aren’t as susceptible to the bugs even in highly populated fluids, but degradation does eventually occur. So a good surface program is important regardless of the polymer you choose.
Hold that thought for a minute. In either case the downhole effects of the biocide added to protect your polymer in the surface can be equally remarkable or devastating on the success of the treatment long-term. Most of the bugs on the surface are aerobic (oxygen loving) in nature and a good biocide program will kill them quickly and easily. But guar and PAM still get pumped down the hole and introduced as a nutrient to an anaerobic (no need for oxygen) population in the reservoir. While the aerobic population would likely die in the anaerobic reducing environment of the reservoir anyway, the starved anaerobic population will now thrive with the thousands of cubic meters of nutrients that have been introduced. Consider that the downhole population usually consists of Sulfate Reducing Bacteria (SRBs), Acid Producing Bacteria (APBs), Iron Reducing Bacteria (IRBs), and others. That’s enough of a list when you consider that the SRBs generate hydrogen sulfide (H2S) which corrodes pipe and the other two corrode pipe directly. Biocides added at the surface have to also protect long-term. Gluteraldehyde and THPS are the typical biocides used in frac jobs and they kill quickly at the surface, but degrade as quickly with little or no lasting effect. Other biocides like DBNPA and quats offer quick kill and a little longer lasting effects, but may have some environmental issues associated with them in certain areas. TDTT offers good long-term kill, but is not particularly good at quick kill on the surface. Fortunately, some of these, like DBNPA or quats and TDTT can be dual injected to achieve both quick kill on the surface and long-term preservation down-hole.
Let’s go back to the surface treatment for a minute and look at the biocide effect directly on the polymer. While we are most concerned with whether biocides kill the bugs and protect the polymers there are also component reactions to consider within the fluid. It can get complicated here as new components, biocides, polymers, and fluid characteristics like pH are being introduced all the time. Gluteraldehyde and THPS, for example, are the most used biocides for surface water treatment, but addition of these products has an adverse effect on both guar and PAM polymer viscosity. Not a huge amount of degradation of course, in comparison to
having no biocide treatment at all, but DBNPA and quats offer quick kill along with TDTT for long-term kill without the degradation to the polymers provided pH and other components like clay stabilizers are also compatible. Since dual addition was mentioned be aware that even certain biocide interactions take place. For example BIT is incompatible with any aldehyde and TDTT particularly in water. At this point it is probably easier to say – make sure your chemistry suppliers are checking compatibility with the full set of chemistries you are trying to use on a well and have a logical alternative plan through the test work that gives you realistic set alternatives if you find serious interactions.
The previous paragraph makes it sound like there are no alternatives; picking the best of the worst to pump downhole. Compatibility testing can also find you some real synergistic improvements in the frac fluids of choice. For example, TDTT works well with certain emulsion PAM packages, offering 5-10% faster viscosity generation (remember ‘time to solution’) and 5-10% higher friction reduction from higher solution viscosity. The ultimate result can be a lower dosage of polymer that still gives higher performance in the end. TDTT also appears to extend oxygen scavenger performance driving ORP lower into the reduction values than other biocides which lower oxidation potential and reduces short-term corrosion and long-term deterioration of tubulars. Again, nothing replaces data run with the actual set of components in the frac fluid compared to a few alternatives.
Conclusions
This is not an exhaustive list as you well know if you’ve done this work already. There are reservoir engineering and mechanical factors to also consider as well as the economics of both the gas being produced and the service being performed. This was only intended as a check list to remind us of some additional considerations in designing and developing frac jobs. One last reminder – nothing can replace data and the continued search for a new way of doing what needs to be done to produce tight gas. The growth of the market itself in the last 10 years is proof of that!