Letters

New correlations
The article "New correlations estimate Pb, FVF" by L.L. Levitan and M. Murtha reports the results of one of the most glaring examples of unscientific and pseudotechnical studies that I have ever seen (OGJ, Mar. 8, 1999, p. 70). I am both surprised and disappointed that OGJ chose to print it.

I have long admired OGJ for presenting the more practical and "hands-on" type of papers, as opposed to those intended for a more academic readership. This particular article, however, has extended the envelope much too far.

Because of the simplicity of the black oil model, correlations based on it tend to be rather more specific to the data used to generate them than applicable to crude oil and gas systems in general. For this reason, almost all such correlations that have been published in the past 10 years or so have contained a specific geographic reference, either in the title of the paper itself or prominently within the text.

As just one example, the title of the Khairy, et al., article to which the above authors refer is "PVT correlations developed for Egyptian crudes" (OGJ, May 4, 1998, p. 114). There are many others that are similarly referenced to diverse areas:

North Sea-Glaso, O., "Generalized Pressure-Volume-Temperature Correlations," JPT, May 1980, p. 785.
Middle East-Al-Marhoun, M.A., "PVT Correlations for Middle East Crude Oils," JPT, May 1988, p. 660.
Western Canada-Asgarpour, S., et al., "Pressure-Volume-Temperature Correlations for Western Canadian Gases and Oils," Journal of Canadian Petroleum Technology, Vol. 28, No. 4, July-August 1989, p. 103.
Gulf of Mexico-Petrosky and Farshad, "Pressure-Volume-Temperature Correlations for Gulf of Mexico Crude Oils," Paper No. SPE 26644, 68th Annual SPE Technical Conference & Exhibition, Dallas, September 1993.
U.A.E.-Dokla, M.E., and Osman, M.E., "Correlations of PVT Properties for U.A.E. Crudes," SPE Formation Evaluation, March 1992, p. 41.

The Standing correlation (Standing, M.B., "A Pressure-Volume-Temperature Correlation for Mixtures of California Oils and Gases," Drilling Production Practice, API, 1947 p. 247) was developed primarily with California crudes, and the Lasater (1958) and the Vasquez and Beggs (Vasquez, M., and Beggs, H.D., "Correlations for Fluid Physical Property Prediction," Paper No. SPE 6719, 52nd SPE Annual Technical Conference & Exhibition, Denver, 1977) correlations were mostly based on data for crudes from the southwestern U.S.

The authors of the OGJ article demonstrate their poor knowledge of the background history of the correlations they discuss and, indeed, of black oil technology in general. Furthermore, their premise that one should develop a new correlation, not using measured data but using values computed by existing correlations, is at best questionable for making a useful contribution to the relevant technology.

As the foundation of this dubious enterprise, they chose the Vasquez and Beggs correlation because "it is the most accurate and reliable." Who, besides them, says so? The authors offer no justification or substantiation for making this bold declaration.

In fact, one of the reasons there are so many geographically based correlations is that the available methods, Vasquez and Beggs included, did not give acceptably accurate results for the crude oil and gas systems in those areas. Many of the papers referred to above give ample comparisons with measured data to demonstrate this fact.

The authors also state that the Standing correlation "is close to the Vasquez and Beggs correlation" but that, when those two procedures showed considerable disagreement, "the Lasater correlation was used for the next step checking." Why? What does "close to" mean?

If the Lasater correlation was to be used as the arbiter of disagreement, why did the OGJ authors not choose it as their basis in the first place? What makes it special in terms of deciding which was right or wrong, particularly since we are only discussing predictions, not comparisons with actual data.

To include subsequently the Khairy correlation in their discussions further underlines their failure to appreciate the limitations that must be applied in order to use black oil model technology intelligently.

The article contains other generalizations and unsubstantiated statements. In the second paragraph, for example, the authors state that the "resulting calculations provided a better description of the experimental data than existing models by Aziz and Orkiszewski." What data? What is their basis for making such a statement?

Both of those procedures are contained in numerous commercial computer programs as part of sophisticated calculation algorithms, for which the PVT description of the fluids is only one part, albeit an important one. How does their implementation compare with generally accepted procedures?

I could go on at some length, but presumably I have made my objections clear. Apart from serving as a bad example, this article makes no useful contribution to fluid property or PVT behavior technology and should never have been published. I hope that it somehow just "slipped through" and does not represent a change in the good judgement that has traditionally been demonstrated by OGJ.

Garry A. Gregory, President
Neotechnology Consultants Ltd.
Calgary
Adjunct Professor, Department of Chemical and Petroleum
Engineering, University of Calgary

Authors' reply
We are pleased that our article has generated such a passionate response. However, we feel that Mr. Gregory is missing the premise of our article.

One of the main critical points raised by Mr. Gregory is that the correlations used in our study are based on simple black oil models that are data-specific and not applicable to general oil and gas systems.

Our study evaluated empirical correlations estimating bubble point pressure (Pb) and formation volume factor (FVF). We selected three correlations: Standing's, Lasatar's, and Vasquez-Beggs's. These correlations were developed by use of experimentally measured Pb obtained from PVT analyses on reservoir fluid samples.

We concentrated on shared parameters (gas/oil ratio, oil and gas relative densities, and temperature). The investigated correlations gave good results over a wide range of oil systems. Hence, we believe that these correlations are applicable to oil and gas systems within the recommended ranges of parameters.

Obviously, using measured data obtained from physical fluid properties is consummate in any study. We never suggested the contrary, as Mr. Gregory says we do. This information, however, is frequently unavailable. Retrieving representative samples is a complex and costly process. Thus, oil scientists and engineers commonly use empirical correlations.

Our premise and contribution is the development of new semi-empirical correlations to estimate Pb and FVF using dimensionless variables. The correlations are uniform at all ranges of parameters and can be easily integrated and differentiated.

Mr. Gregory emphatically expressed disapproval in choosing the Vasquez and Beggs correlations as the foundation of our study. We placed significant weight on the Vasquez-Beggs experimental data because they contained more than 5,000 measured points with a mean error of -0.7%. The data encompassed very wide ranges of pressure, temperature, oil gravity, and gas gravity.

Lastly, Mr. Gregory claims that we made generalizations and unsubstantiated statements with regard to our comparison of our commercial software program to existing models by Aziz and Orkiszewski.

We ran experimentally measured data presented by Aziz using our well bore hydraulic program. We compared the calculated bottom hole pressures to experimental data presented by Aziz, and our program generated a lower deviation.

We realize that our findings need to be tested further and invite added investigation. Perhaps Mr. Gregory may be interested in trying out our well bore hydraulic program that incorporates our general correlations. He may be pleased with the results.

Leonid Levitan

Industry economics changing
The U.S. competes in a global economy dependent on energy use, yet the most successful and energy-dependent nation on earth has not consistently implemented a coherent energy policy. We need to implement efforts to see that our energy policies improve opportunities for the domestic oil and gas industry and reduce our dependence on foreign oil imports. Energy consumption has doubled in the past 25 years, and fossil fuels continue to account for 85% of the energy consumed in the U.S.

Over the past decade, technological innovation has rapidly emerged as one of the critical success factors in the oil and gas industry. The industry is one of the most adaptable business creations of the industrial age. From its earliest days, it has been the victim and beneficiary of market fluctuations larger than those that have confused and defeated the leaders of other industries. Yet it has not only survived but thrived.

New technology in every part of the business drove down costs so that profits could be protected at crude prices as low as $11/bbl. Previous wisdom was that the industry couldn't make money at less than $23/bbl. Several major breakthroughs have provided performance improvements that have changed the economics of the business. However, continued innovative technology will be needed, even as oil prices recover, because there will be hesitant reservations that prices will stay high enough to merit future investments over the life of the asset.

In the past, when energy companies had large staffs, they acted as the central repository of information and jealously guarded the information that gave them advantages. They generally limited outside access to that information in order to protect themselves or their turf. With the ongoing cuts of most of these staff people through cost-cutting initiatives, access to propriety information has in most cases been lost. Loyalty is a thing of the past.

So many cuts have been made in continuous downsizing, there is no longer available the expertise to access the information available. The result is that historical solutions are getting reinvented every day in the field and performance is being sub-optimized. Using outside engineering and contracting services is providing the expertise to fill this worldwide growing void. It enables the oil and gas companies to add depth without having to recreate central bureaucratic in-house departments.

Contractors are now having to furnish the expertise on an equal basis to all the customers. All data is shared to solve problems in locations around the world. Collectively, the industry shares the expertise and experience when and where it's needed. However, while the industry continues to make itself more efficient, cleaner, safer, and smarter, our nation's oil and gas resource base continues to shrink. As this takes place, access restrictions and higher geologic and economic risk at home help make investments in other countries' oil and gas sectors look more attractive.

Our history of offshore drilling shows that we can manage so that oil and gas drilling and production are attractive to investment and competitive with other world oil and gas prospects. Make no mistake: The Gulf of Mexico is competing with basins around the world for development capital. Our government must create and sustain a stable, consistent business environment and help, rather than hurt, which is instrumental in reversing the decline in U.S. oil and gas production rates. The U.S. deepwater sector needs and deserves both the continued royalty relief and stable, consistent lease terms.

Dailey J. Berard
New Iberia, La.

Egyptian product pipeline
This is in reference to the article "Pilot line verifies calculations for interface length, mixing" (OGJ, May 24, 1999, p. 66). My observations are as follows:

While describing the Suez-to-Mostorod pipeline, the authors specify 10-in. OD and 12-in. OD sections. In calculations, however, they use these numbers as IDs.It would have been better had the authors described the size of sections as nominal sizes and specified the thickness of the respective sections, so that it is easier for the reader to correlate the calculations.
The authors have very easily summed the interface volumes while performing the validity check of the existing equations. In actuality, however, we cannot sum the interface volumes of two different sections. Here are some calculations for benzene-LPG interface to show this.
A. Case as given in the article:
Section detail Interface volume, cu m
10-in. ID, 89 km 23.69
12-in. ID, 51 km 28.93
Total 52.62
B. Now let us assume that the first section is divided into two parts:
Section detail Interface volume,cu m
10-in. ID, 44 km 16.66
10-in. ID, 45 km 16.85
12-in. ID, 51 km 28.93
Total 62.44
C. Let us take one more case in which the second section is also divided into two parts:
Section detail Interface volume,cu m
10-in. ID, 44 km 16.66
10-in. ID, 45 km 16.85
12-in. ID, 25 km 20.17
12-in. ID, 26 km 20.57
Total 74.25
It can be observed that we obtain 52.62, 62.44, and 74.25 cu m of interface volume for the same length of the pipeline, i.e. 140 km. It is very clear from this that we cannot simply sum the interface volumes of different sections.
It is unclear from the article how anyone can predict the density variation with time without considering the flow rate in the pipeline and the densities of two different products.
The problem that engineers generally face during designing a pipeline is how to predict the density variation within the interface's length.
Fitting the experimental data of density and time can be useful for only one system. Either it should have been generalized for any system or the subject should not have been touched upon in this article.
Finally, the authors have mentioned that they have taken the equation from "Institute of Mechanical Engineering" but give no reference.
Providing the reference of the equation would have been very helpful, because then the reader could go through the original paper and discover what parameters were considered for development of this equation.