Method for measuring physical characteristics in a pipeline without tapping

Abstract

A fluid-filled pipe is investigated by a non-intrusive procedure to determine whether the fluid is liquid or gaseous, whether it is flowing or static, the direction of flow if it is flowing, the approximate pressure if it is a gas, and the rate of flow if it is a gas. The results are obtained by applying a heater to the surface of the pipeline and measuring the upstream and downstream temperatures of the surface of the pipe wall before and after the beginning of the application of heat. Some of the data generated is compared with data from another source to assist in the determination of the characteristics to be ascertained.

Description

FIELD OF THE INVENTION

This invention relates to a method and apparatus for determining physical characteristics of a fluid inside a pipeline without breaching the surface of the pipeline.

BACKGROUND OF THE INVENTION

This invention came about as a result of public utilities' need to know the characteristics of the contents of pipelines which may be excavated by repair crews.

Various techniques have been used by public utilities to provide later identification of specific pipe systems after they are installed and buried. The problem is, nature is not particularly cooperative in maintaining the schemes of identification originated by man.

Survey crews and draftsmen are often quite specific in preparing crossing locations for buried conduits and they are more than adequately related to landmarks available at the time the pipe is buried. Unfortunately, the street curbline may be changed by the city, the centerline of the street may be changed by the government entity responsible for maintenance, the alignment of power lines between power poles may be shifted, other pipelines or electrical conduits may be buried above or below the critical pipeline or installed at an angle crossing the same.

The problem created is one to be solved by an excavation crew on-site. It is critical for the safety of working crews and citizens near the site of an excavation for the operating crew to know the characteristics of the contents of pipelines.

For example, an excavation crew from a natural gas public utility working in an urban area will believe that it knows about where a particular pipeline is located. This information will be based on data collected from the engineering department of the utility in question. The problem is that the on-site excavation crew may find the wrong pipe and not know it. The potential for disaster is self-evident if the crew thinks that the excavated pipe is a water pipe and it turns out to be a natural gas transmission line operating at 150 psi and flowing at 25 ft/sec.

There is a need in the utility industry for an apparatus and procedure for determining physical characteristics of the fluid inside a pipeline without breaching the sidewall of the pipe. Such data would ensure that the excavation crew is operating on the right pipeline and in addition will take the necessary precautions for safety of the population when the pressure flow rate and direction of flow are available prior to the time the intended structural modifications are made to the pipeline.

SUMMARY OF THE INVENTION

Certain generalities and experimental data referred to subsequently come from experimental data involving carbon steel pipes which meet all the requirements of API 5L Schedule 40 specifications. Such generalities involve pipes ranging from 2 to 12 inches in diameter and having a sidewall thickness in the range from 0.156 to 0.375 inches. These are standard thicknesses. It will be quite obvious that the inventive concepts defined herein are applicable to a wider range of sidewall thicknesses and diameters than the standard thicknesses and diameters enumerated above and specified in industry standards.

A technique for achieving the desired result of determining physical characteristics of the fluid content of a just-excavated pipeline is accomplished by heat-transfer characteristics of various fluids at diverse pressures and under static or flowing conditions.

The apparatus of this invention involves a heater applied to the surface of the pipeline combined with a plurality of temperature-sensing devices located at the site of the heat application to the pipe surface and various locations axially transverse to the location for the application of heat.

The general inventive characteristics are applicable both to a ring-type heater extending completely around the circumference of the excavated pipe and a patch-type heater applied to the surface of the pipe which extends less than the full circumference of the pipe. It is anticipated that the patch-type heater will be the preferred apparatus for use in the field because it is smaller in size and less cumbersome to use.

The basic concept involves mounting the heater and the transversely located sensing devices on the pipeline, measuring the temperature of the pipeline surface and then applying heat at the heat source while measuring the temperature changes at the site of the heat source and the transverse locations. The pattern of heat transfer or temperature pattern allows the evaluation of the physical characteristics of the contents of the pipeline.

The ways to interpret the data generated will be described in more detail subsequently. Objects of the invention not understood from the above will become clear upon a review of the drawings and the description of the preferred embodiments which follow.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a fragmentary schematic view of an excavated pipeline with the apparatus of this invention mounted in operative position;

FIG. 2 is a perspective view of the open clam shell shaped housing of this invention;

FIG. 3 is a perspective view of the clam shell housing of FIG. 2 closed around a pipe;

FIG. 4 is an end elevational view of the open clam shell housing of FIG. 2, partially in phantom;

FIG. 5 is a chart of sidewall temperatures for natural gas and water-filled pipes with a flow velocity of 6 ft/sec. and a pressure of 140 psia;

FIG. 6 is a chart of temperature differentials between the top and bottom of a pipe under the same conditions as the fluids in FIG. 5;

FIG. 7 is a chart of calculated, projected sidewall temperatures of a pipe with static gaseous contents at various pressures;

FIG. 8 is a chart of calculated, projected top-to-bottom temperature differences of a pipe with static gaseous contents at various pressures;

FIG. 9 is a chart of observed experimental temperature rise with a ring heater in a pipe with static gaseous contents versus time at a location about 5 inches from the centerline of the 250 W heat source;

FIG. 10 is a chart providing an alternative plot of the same data as FIG. 9 but with the axes defining temperature rise versus pressure;

FIG. 11 is similar to FIG. 9 except that it is a plot of observed experimental data with a heater input of 250 W taken at a point 6 inches from the centerline of the heater;

FIG. 12 is a chart similar to FIG. 10 except that it is an alternative plot of patch heater experimental data with a heater input of 250 W and readings taken 6 inches from the centerline of the heater;

FIG. 13 is a chart showing calculated and experimental data of a plot of top-to-bottom temperature differences versus pressure with flowing gas at various velocities with temperature reading 10 inches downstream from the centerline of a ring heater;

FIG. 14 is a chart showing calculated and experimental data of a plot of the sidewall temperature differences versus pressure for various flow rates for gas with temperature readings 10 inches downstream from the centerline of the heater;

FIG. 15 is a chart of the rate of rise of top-to-bottom temperature difference versus time in a 4 inch pipe with readings taken at 6 inches downstream of the centerline of a patch heater for various pressures with gas flowing at a velocity of 4 ft/sec.;

FIG. 16 is a chart of the rate of rise of temperature of a patch heater measured at the sidewall of a pipe at a point 6 inches downstream from the centerline of the heater, plotted for various pressures with a gas flow rate of 4 ft/sec. versus time;

FIG. 17 is a plot of estimated versus actual pressures with FIGS. 15 and 16 parameters;

FIG. 18 is a plot of estimated versus actual velocities with FIGS. 15 and 16 parameters;

FIG. 19 is a plot of estimated pressure vs. measured pressure for a static gas using a 2-X linear fit for Schedule 20;

FIG. 20 is a plot of estimated pressure vs. measured pressure for static gas using a 4-X linear fit for Schedule 20;

FIG. 21 is a plot of estimated pressure vs. measured pressure for static gas using a 2-X logarithmic fit for Schedule 20;

FIG. 22 is a plot of estimated pressure vs. measured pressure for static gas using a 4-X logarithmic fit for Schedule 20;

FIG. 23 is a plot of estimated pressure vs. measured pressure for static gas using a 2-X linear fit for Schedule 40;

FIG. 24 is a plot of estimated pressure vs. measured pressure for static gas using a 4-X linear fit for Schedule 40;

FIG. 25 is a plot of estimated pressure vs. measured pressure for static gas using a 2-X logarithmic fit for Schedule 40;

FIG. 26 is a plot of estimated pressure vs. measured pressure for static gas using a 4-X logarithmic fit for Schedule 40;

FIG. 27 is a plot of estimated pressure vs. measured pressure for static pressure using a 2-X linear fit for Schedule 80;

FIG. 28 is a plot of estimated pressure vs. measured pressure for static gas using a 4-X linear fit for Schedule 80;

FIG. 29 is a plot of estimated pressure vs. measured pressure for static gas using a 2-X logarithmic fit for Schedule 80;

FIG. 30 is a plot of estimated pressure vs. measured pressure for a static gas using a 4-X logarithmic fit for Schedule 80;

FIG. 31 is a plot of regression coefficients as a function of pipe wall thickness;

FIG. 32 is a plot of estimated velocity vs. measured velocity for Schedule 40 pipe; and

FIG. 33 is a plot of estimated pressure vs. measured pressure for Schedule 40 pipe.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

For purposes of this invention, the broad concepts are to be stated relatively simply because, after the inventive concepts were conceived and the experimental data generated to prove the same, the inventive concepts proved to be relatively straight-forward.

At the outset, a pipe 10 is first located and a section is excavated as shown in FIG. 1. The excavation crew hopes it knows what is inside the pipe, but its collective intuition must be verified with apparatus and procedural steps to confirm the intuitive conclusion.

The pipe 10 is excavated to expose at least three lineal feet of pipe to accommodate a clam shell shaped housing 34 which is about two feet long. At least six inches of soil is excavated below pipe 10 to allow proper installation of the testing equipment.

Normally, buried pipes are coated with an asphalt composition or a wrapping to minimize exposure to water and corrosive gases and liquids. To allow proper equipment operation, the coating should be removed for a least eighteen lineal inches or at least where the heater and temperature sensors of this invention contact the pipe surface. The exposed surface should be cleaned, preferably with a power wire brush, wherever the temperature sensor and heater are to contact the pipe.

An ultrasonic thickness gage may be used to determine pipe thickness as will be explained subsequently. The outside diameter of the pipe is input to the computer program used in this invention and if a thickness gage is used its measurement is also input.

A heater 12 is applied to the surface of the pipe using the clam shell structure 34 illustrated in FIGS. 2-4. FIG. 1 illustrates the heater 12 as being a patch extending considerably less than completely around the circumference of the pipe. A heater of larger proportions could be used including a completely annular heater extending completely around the pipe. Experiments have indicated that a rectangular patch heater one inch by two inches is adequate. Preferably the sensors 14-24 are K-type, 1/16-inch thermocouples, but numerous other sensors including other thermocouples are suitable.

The first determination of physical characteristics of the fluid inside the pipe may be made relatively quickly. This is achieved by applying temperature sensors 14, 16, 18, 20, 22 and 24 to the surface of the pipe at locations axially transverse of heater 12.

It is preferred that the pipe axis 26 be relatively horizontal for purposes of this invention for reasons which may be understood from the following description. But at the outset it should be recognized that the inventive concept is based on heat transfer characteristics of both fluids and solids and it also involves heat transfer by conduction, convection and radiation under static, laminar and turbulent flow conditions. Where the axis 26 deviates to some great degree from horizontal, the convection cells generated inside the pipeline may distort the temperature patterns and heat transfer data because of gravity, among other things.

It is preferred that the vertically aligned locations of sensing devices 14, 16, 18 and 20, 22, 24 be axially spaced from the heat source 12 by from 2 to 25 inches and most preferably from 3 to 8 inches.

The first physical characteristic determination to be made is whether the fluid inside pipe 10 is liquid or gaseous. This is accomplished by measuring the temperature of the sidewall at, for example, sensor 16 before controller 28 activates the power source 30 to begin the generation of heat by heater 12. The magnitude of heat from power source 30 to heater 12 is controlled by controller 28 through power meter 32. Then the heater is activated. The standard wall thicknesses for pipes of given diameter are a part of the computer program used to control the controller 28 unless a determination has been made on site, in which case the measured wall thickness is input to the program.

After a short period of time, preferably 3 to 30 minutes and most preferably about 5-6 minutes, a second reading is taken by temperature device 16. Where there is no substantial change in temperature between the first reading and the second reading at device 16, one can conclude that pipe 10 contains a liquid. This conclusion may be drawn from the fact that a comparatively dense liquid is a sufficient heat sink as to transfer almost all of the heat from heater 12 to the liquid without a transverse longitudinal flow along the pipe sidewall. The distance from heater 12 to temperature sensor 16 is 50 or 100 times greater than the thickness of the pipe wall to the liquid. On the other hand, if within the specified period of time, the temperature at sensor 16 has increased, it may be concluded that the contents of the pipe is a gas. This conclusion may be drawn from the fact that a comparatively less dense gas is unable, because of its density and low thermal conductivity, to extract all of the heat from the pipe wall generated by the heater 12.

Consider another application which could tell the crew how much liquid is in the pipe. First measure the temperatures at 20, 25 22 and 24 (they should be about the same). Then apply heat at 12.

After a period of time, again measure the temperatures at 20, 22 and 24. If the first temperature measurement at 20 is about the same as the second temperature measurement at 20, one can conclude that there is liquid in the pipe. If the first temperature measurement at 22 is about the same as the second temperature measurement at 22, one can conclude that the pipe is at least half full of liquid; if the second temperature measurement at 22 is much higher than the first temperature measurement at 22 and both measurements at 20 are about the same, one can conclude that the pipe is less than half full of liquid. If the second temperature measurement at 24 is about the same as the first measurement at 24, one can conclude that the pipe is almost full of liquid; if the second temperature measurement at 24 is much greater than the first temperature measurement at 24 and both measurements at 22 are about the same, one can conclude that the pipe is more than half full of liquid, but not completely full.

The next consideration is to determine whether or not there is fluid flow within pipe 10 and if so, what direction the fluid is flowing. Indeed, the data generated early on will probably be simultaneous with the determination of the liquid-gas issue. It involves taking temperature readings upstream and downstream of A heater 12. For example, temperature sensors 16 and 22 mounted on the sidewall about 90.degree. down from the top of the pipe are located such that a line drawn between sensor 16 and sensor 22 is approximately parallel with pipe axis 26. That line passes from one sensor through the heat source to the other sensor. Sensor 16 and sensor 22 are spaced equidistant from heater 12.

An observation of FIGS. 5 and 7 will help illustrate the conclusions which may be drawn from data generated from sensors 16 and 22. It is intuitive that a temperature pattern of liquid or gas in a pipeline heated according to FIG. 7 will have a plotted temperature pattern defining a classic bell-shaped curve. The bell-shaped curve for the liquid as seen in FIG. 5 is much flatter than the one for gas because of the greater ability of the dense liquid to absorb heat. Accordingly, a proper reading after a short period of time, both upstream and downstream of heater 12, confirms whether the contents are liquid or gaseous. If the contents are liquid and static, the temperatures at 16 and 22 should read about the same and not greatly different from the temperature reading at the heat source, see FIG. 5. On the other hand, where the contents are gaseous, the temperature at heater 12 is substantially greater then the temperature readings at 16 and 22 and the readings at 16 and 22 are elevated as compared to the initial reading before the controller 28 is activated.

FIG. 5 is a plot of temperatures taken at sidewall locations about 90.degree. down the side of the pipe at locations 5, 10 and 15 inches upstream and downstream of heater 12. In the illustrated curves, the pipe was 6 inches in diameter with fluid contents having a velocity of 6 ft/sec. and a pressure of 140 psia for both natural gas and water. The temperatures plotted are at steady state after 11/2 hours of heating.

A couple of things should be mentioned at this point which could have an effect on some of the readings. The locations of temperature sensors 14 and 24 are at the top of the pipe. Readings only at the top of a pipe may be misleading because pipes do not run full of liquid as a general rule. There may be a small trough of air at the top of a water-filled pipe which could give misleading data.

A reading of the temperatures at the sidewall locations 16, 22 and an observation of FIG. 5 will show the direction of flow of the flowing gas. Observe that the temperature pattern of the natural gas is slightly skewed to the right indicating that the gas is flowing from left to right. The temperatures at 5, 10 and 15 inches to the right of the heater are slightly higher than the temperatures of the gas 5, 10 and 15 inches to the left of the heater. The pattern of temperatures for water is not so clear-cut because of its greater thermal conductivity.

These data have established whether the pipe is filled with gas or liquid and whether it is static or flowing if it is gas. Further, the data allows one to determine the direction of flow of the flowing gas, if it is flowing. This information may be enough for the excavation crew to determine that they have not located the desired pipe. If that is the case, it is very important.

An observation of FIG. 6 shows an interesting technique useful for gaseous flow, but of no significance at all to liquid flow. It is that a reading of temperature differentials between the top and bottom of the pipe can give some very useful information. The curve in FIG. 6 is for natural gas and it is shows the temperature difference between readings of the sensors at the top of the pipe, at 14 and 24, as compared to readings at the bottom of the pipe, at 18 and 20. These data in FIG. 6 were generated from natural gas flowing in a 6 inch diameter pipe at a velocity of 6 ft/sec. and at a pressure of 140 psia.

It will also be observed in FIG. 1 that an insulated clamshell device or jacket 34 surrounds both the pipe 10 and the apparatus mounted on the pipe wall (thermocouples, heater, etc.). This enclosure preserves the heat generated by heater 12. Insulation of and screening from the environment is intended to mitigate against temperature differentials between the pipe excavated in Wisconsin in January and one excavated in the Mojave Desert in July. The intent is to have the heat generated by heater 12 as the only heat generating device in the test area. Clearly the end pieces 38 must be replaced for each diameter of pipe.

FIG. 7 illustrates the symmetric steady state profile of different pressures plotted for distance from the center of the heater after a period of about 90 minutes or longer. With this information available for comparison based on a reading at a location transverse to the heater, one can predict the pressure of the static gas inside a pipe because the temperature and distance are known to the excavation crew. The chart of FIG. 7 may be displayed on the computer screen and the gas pressure can be scaled by the operator. Similar information can be derived from the chart illustrated in FIG. 8. FIGS. 7 and 8 are based on static conditions in relatively ideal conditions and in field operations the excavation crew does not have enough time to allow the excavated pipe to come to steady state conditions. Accordingly, an alternative set of parameters is illustrated in FIGS. 9 and 10 which may be used in the field more readily. The data illustrated in FIG. 9 is intuitive in that the fastest temperature rise is at the lowest (16 psia) pressure and the slowest rise is at the highest (152 psia) pressure. It is intuitive because it is to be expected that greater amounts of heat will transfer to the denser gas at the higher pressure. Accordingly, the temperature at the pipe wall does not rise as quickly at the higher pressure.

FIG. 10 shows six curves plotted at 30, 40, 50, 60, 70 and 80 minutes after the onset of temperature rise at a location 10 inches from the centerline of the heater 12. This chart allows an estimate of the gas pressure because the excavation crew will know the period of time and the temperature. This can be translated into an estimate of pressure using FIG. 10 displayed on the computer screen for whatever size pipe was execavated. Unfortunately, the time involved is burdensome. Note that the curve at 30 minutes is essentially a flat curve making it very difficult to give a reasonable estimate as to the internal gas pressure. On the other hand, at 80 minutes the curve is much better because of its change in slope as pressure increases.

FIG. 9 shows the temperature rise in the sidewall of the pipe measured 5 inches downstream from a ring heater with a 250 W heat input and with flowing gas.

FIG. 11 has the same location for the heater as in FIG. 9 with slightly different pressures using a patch heater with a 250 W heat input and measured 6 inches downstream from the center of heater 12 and with static gas.

Both systems as illustrated in FIGS. 9 and 11 were operating at 250 W. The ring heater distributes the heat over a wider area so that the heat flux and overall temperatures are much lower under the heater compared with the patch heater which delivers the same power to a much smaller area. The local higher temperatures might set up stronger convection currents inside the pipe so that the temperatures rises faster for the patch heater system.

The heat flow method relies on the normal response of the pipe wall to a heat input. The peak wall temperature occurs directly under the heater. For the purpose of estimating pressure, the heat input should be as high as possible, which implies the wall temperature under the heater will be very high, but there is an upper limit to the temperature determined by the metallurgy of the pipe. From other data it has been determined that a safe upper limit is about 800.degree. F. Accordingly, the heat sensor incorporated as at heater 12 combined with controller 28 keeps the pipe wall lower than 800.degree. F.

While the discussion to this point has referred to pipe size where nominal diameters and standard sidewall thicknesses are known, sometimes the pipe wall thickness may be unknown. An ultrasonic wave generator may be used to determine the thickness if desired. Such devices are well known in the industry and need not be described in detail. One such hand held device is marketed by Sandtec Products of Fairfield, Ohio under the Trademark T-MIKE-E. Such a wall thickness measuring unit 40 is illustrated schematically in FIG. 1.

It is intuitive that a thick pipe wall will absorb more heat than a thin wall. Further, heat flows faster in solid metal than in gas. Clearly the wall thickness affects the speed of transverse heat transfer. Therefore, determining the wall thickness prior to, or during, the application of heat may allow a more accurate estimate of pressure.

Where a sidewall thickness measuring apparatus is available, the data is input to the computer along with the measured outside diameter of the pipe. The outside pipe diameter is input in any case and the estimates of pressure, flow rate, etc. may be achieved with or without a measured sidewall thickness. Data input including temperatures are then correlated with curves in the data base of the program to achieve the estimates of pressure, flow rate, etc.

The data from FIGS. 9 and 11 are used to produce the curves shown in FIGS. 10 and 12. The curves of FIGS. 10 and 12 are plots of temperature rise on the pipe surface as a function of internal gas pressure and with flowing gas. As seen in FIG. 12, the slopes of the curves are greater for the patch heater than for the ring heater as illustrated in FIG. 10. The 30 minute curve from the patch heater has a reasonable slope to permit an estimate of pressure to be made. The slope of the curve for 60 minutes is even steeper. Therefore, an operator at the excavation site can obtain a rough estimate at 30 minutes and perhaps elect to wait another 30 minutes for a heater based on curves in FIG. 10. An operator would have to wait at least 50 minutes to get even a reasonably rough estimate from the ring heater system. Therefore, the patch heater does provide a more effective and accurate system while improving the field applicability.

FIG. 13 is more helpful in this context. It is a plot of temperature difference as a function of pressure at 10 inches downstream of the center line of a ring heater having a 250 W heat input. There is a fairly good correlation between the numerical analysis (num) and the experimental results actually measured (exp) and it is significant because the correlation has so little to do with pressure and it is controlled almost completely by velocity. It was intuitively expected that the top-to-bottom temperature differences would reflect pressure differences due to the varying strength of the free convention cells generated in the pipe during the heating process. However, both the numerical predictions and the experimental results show that the main influence on the top-to-bottom temperature difference is from velocity. Thus, the top-to-bottom temperature difference is a good indicator of velocity and the downstream top-to-bottom temperature measurement yields a good estimate of gas velocity which is nearly independent of pressure.

The data being discussed at this point relates to the ring heater, but it may be freely extrapolated to the patch heater for purposes of this discussion.

Once the approximate velocity of the gas is known (from FIG. 13), the pipe sidewall temperature proves to be indicative as can be ascertained from the data in FIG. 14. Those data show that there is excellent agreement between numerical predictions and experimental data. They also show that the temperature measurement is more sensitive at low velocities. That is the slope is steeper, which is expected and advantageous for differentiating between low pressure ranges.

The experimental and numerical results of FIG. 14 both demonstrate that a pressure estimate can be obtained from the top-to-bottom temperature difference and the sidewall temperature. First, the temperature difference is used to obtain an estimate of velocity from FIG. 13. Then, the data from FIG. 13 using the velocity is used in FIG. 14 to determine the correlation between sidewall temperature 10 inches downstream from the centerline of the heater and the internal pressure read on the X axis of FIG. 14. The measured sidewall temperature is used to estimate the pressure.

A major goal of this invention is to estimate pressure within 30 minutes under flowing gas conditions. Unfortunately, as illustrated in FIG. 12, the period of time to reach near steady state exceeds this 30 minute time period.

Because the patch heater proved to be effective for estimating pressures of static gas, the patch heater has been tested under flowing gas conditions. It was necessary to get some estimate of pressure and velocity within a 30 minute period without waiting for steady state temperatures. Accordingly, work was done to obtain an estimate using the temperature difference between the top and bottom of the pipe downstream of the heater.

FIG. 15 is a plot of the rate of temperature rise measured in a 4 inch pipe 6 inches downstream of the patch heater (dTz denotes the top-to-bottom temperature difference downstream of the heater). The rate of rise of both dTz and the steady state value of the sidewall temperature downstream of the heater at the same location for all the cases of pressure are plotted as a function of time in FIGS. 15 and 16 for a nominal gas velocity of 4 ft/sec. Note that the peaks in the rate occur in under 30 minutes. The behavior of the peaks with respect to time appear to follow power laws with pressure, P, and velocity, V. Thus, the data for heights were modeled with a relatively simple relation given by

H=CP.sup.a V.sup.b (Eq. 1) where a, b, C are parameters that are determined by the data.

The form of the relation is particularly useful because the estimates of the constants can be found using a multiple linear regression by taking the logarithms of both sides. That is

a log(P)+b log(V)+log(C)=log(H) (Eq. 2)

where log(P), log(V), and log (H) are now the X, Y, and Z variables, a and b are slopes, and log(C) is the intercept. Thus, a well-known commercial spreadsheet computer program can be used to obtain the estimates of a, b, and C and need not be discussed in detail.

Two sets of data were used to obtain estimates of pressure and velocity. The first set did not include the static cases, and the second set did. For static gas the velocity equals zero, and the logarithmic approach breaks down. Therefore, when the regression as applied to the set with the static data, V was replaced with V +1. The offset term of 1 was arbitrarily chosen, but was deemed appropriate since the lowest value of flowing gas used in the experiments was one ft/sec. If this power law is kept for the engineering evaluation unit, then the offset term could also be chosen with an optimization approach.

The results of the regression analysis are the estimates of the parameters, a, b and C. The power laws are manipulated to yield estimates of pressure and velocity given the peaks heights or the rates af rise using the estimates of the parameters. The results for the data set without the static cases are shown in FIGS. 17 and 18. They are plots of the estimates of pressure and velocity versus the measured values of pressure and velocity. As seen in these figures, acceptable performance is obtained for both the estimates of velocity and pressure.

Power law regressions to obtain estimates in under 30 minutes are as follows:

Pressure Estimate Without V=0 Data P=2875H.sub.h.sup.-3.278 H.sub..DELTA.T.sbsb.Z.sup.0.9277 (Eq. 3) Pressure Estimate With V=0 Data P=6310H.sub.h.sup.-1.524 H.sub..DELTA.T.sbsb.Z.sup.1.318 (Eq. 4) Velocity Estimate Without V=0 Data and Without Pressure Estimate V=0.1177H.sub.66 T.sbsb.z.sup.-2.700 (Eq. 5) Velocity Estimate Without V=0 Data and Using Pressure Estimate V=0.07481H.sub.66 T.sbsb.z.sup.-2.522 P.sup.0.1705 (Eq. 6) Velocity Estimate With V=0 Data Without Pressure Estimate V=0.4170H.sub.66 T.sbsb.z.sup.-1.982 -1 (Eq. 7) Velocity Estimate With V=0 Data and Using Pressure Estimate V=0.1748H.sub.66 T.sbsb.z.sup.-1.321 P.sup.0.2500 -1 (Eq. 8)

It has been shown that the heat flow method works with both a ring heater and a patch heater. The patch heater is expected to be much easier to use in the field so it is the preferred implementation. Both heaters are resistive heaters. Direct contact heaters are preferred at this time because of their ease of use and inexpensive cost as opposed to induction heaters or radiant heaters. In the experiments indicated above, thermocouples were used as temperature sensors but any temperature-measuring device may be used. An observation of FIG. 1 will show that all of the temperature-sensing devices, including the one directly under patch heater 12, are directly connected to the data acquisition system and computer 36 used by the on-site excavation crew to make the determination on whether or not they have excavated the correct pipe. The data display associated with the system 36 may be in the form of numbers, charts, or whatever works.

One function of the temperature-sensing device beneath patch 12 is to ensure that the temperature does not exceed 800.degree. F. where the metallurgy of the steel pipe could be irreversibly changed.

The heat flow method requires that the heater 12 be given a fixed amount of power for a given test. The computer 36 selects the power level for the size of the pipe being evaluated, whether or not the sidewall thickness is known, and uses this level as the set point for the controller 28. Its output controls and SCR circuit 30 delivers power to the heater. A power meter 32 monitors the electrical power and feeds the value back to the controller 28. FIG. 1 illustrates the controller 28 as an individual unit. It could be implemented in the computer software and interfaced with the hardware through the data acquisition system 36.

Performance of the apparatus tested over a wide range of static and flow conditions (see Tables 1, 2 and 3) to both verify that the device worked adequately and to provide data for the correlations described as follows.

TABLE 1______________________________________Static Field Test Results for Nominal 4-inch Diameter PipeTest Run Pressure, psiaID Actual Measured______________________________________Static Runs, Schedule 20S20-01 15 18.9S20-02 15 24.1S20-03 25 22.8S20-04 25 25.4S20-05 25 19.8S20-06 60 72.5S20-07 60 79.0S20-08 60 58.4S20-09 60 46.7S20-10 100 109.4S20-11 150 128.0S20-12 100 111.2S20-13 100 95.7S20-14 150 137.2S20-15 150 143.0S20-16 100 114.4S20-17 25 26.8Static Runs, Schedule 40S40-01 15 14.2S40-02 15 21.7S40-03 15 22.6S40-04 25 29.4S40-05 25 26.2S40-06 25 27.2S40-07 60 44.2S40-08 60 57.1S40-09 60 53.7S40-10 100 110.5S40-11 100 111.7S40-12 100 99.7S40-13 150 125.3S40-14 150 152.9S40-15 150 135.4S40-16 15 18.9S40-17 15 17.2S40-18 25 20.7S40-19 25 20.3S40-20 60 63.4S40-21 60 65.4S40-22 100 114.0S40-23 100 92.1S40-24 150 135.2S40-25 150 131.5S40-26 100 103.4S40-27 25 20.6S40-28 25 26.8Static Runs, Schedule 80S80-01 15 23.2S80-02 25 29.7S80-03 100 99.5S80-04 60 49.7S80-05 150 152.8S80-06 150 143.5S80-07 25 27.0S80-08 25 36.5S80-09 25 25.6S80-10 60 71.6S80-11 60 57.8S80-12 100 95.2S80-13 100 91.7S80-14 100 116.4S80-15 150 143.8S80-16 150 152.0S80-17 150 122.1S80-18 150 117.0S80-19 150 164.4S80-20 100 102.3S80-21 100 112.1S80-22 100 113.3S80-23 100 114.5S80-24.sup.a 60 135.8______________________________________ TABLE 2______________________________________Flow Field Test Results for Nominal 4-Inch Diameter Pipe,Schedule 40Pressure, psia Velocity, ft/secTest Run Nomi- Nomi-ID nal Actual Measured nal Actual Measured______________________________________Flow Runs, Schedule 40F40-01 150.0 150.4 114.6 15.0 15.0 12.6F40-02 100.00 15.2 12.5F40-03 100.10 11.7F40-04 60.70 10.2 9.9F40-05 100.70 1.0F40-06.sup.b 25.3 1.9F40-07.sup.a 60.0 1.0F40-08 25.30 6.4F40-09 25.00 2.8F40-10 15.60 3.7F40-11 15.50 1.5F40-12 60.40 3.0F40-13.sup.c 41.0 5.6F40-14 80.30 4.9F40-15.sup.a 40.5 8.9F40-16 40.80 1.8F40-17 80.80 1.8F40-18.sup.d 150.0 5.6F40-19.sup.d 60.3 1.3F40-20.sup.d 149.4 4.8F40-21.sup.a 25.5 15.0 9.9F40-22.sup.a 150.1 1.7F40-23.sup.a 149.8 10.1 11.8F40-24.sup.a 60.1 15.2 10.9______________________________________ TABLE 3______________________________________Flow Field Test Results for Nominal 4-Inch Diameter Pipe,Schedule 40Pressure, psia Velocity, ft/secTest Run Nomi- Nomi-ID Actual Measured nal Actual Measured______________________________________Flow Runs, Schedule 80F80-01 150.0 150.2 141.1 16.6 16.7 14.1F80-02 106.0 13.7F80-03 100.4 109.3 16.3F80-04 150.0 152.1 14.3F80-05 100.4 54.9 15.5F80-06 32.5 4.1F80-07 150.1 59.9 2.6F80-08 32.7 2.4F80-09 26.1 2.4F80-10 41.0 11.2F80-11 43.8 2.7F80-12 30.0 3.1F80-13 28.4 2.1F80-14 29.9 5.64.9F80-15 61.6 3.2F80-16 58.4 1.9F80-17 48.7 2.2F80-18.sup.a 135.7 12.1F80-19.sup.f 100.5 57.2 2.6F80-20.sup.g 26.3 2.1F80-21.sup.c 157.0 2.7F80-22.sup.a 151.5 138.9 13.2F80-23.sup.c 149.4 51.7 11.6F80-24.sup.d 10.1 3.3______________________________________ Notes regarding data: .sup.a anomalous .sup.b ended abruptly .sup.c rain interference .sup.d equipment seal failure .sup.e ended early .sup.f hightemperature cutoff .sup.g problem noted

A total of 69 static tests were run to provide the data for the correlations. Two sets of three different regressions have been determined for each pipe wall thickness for a total of 18 correlations. One set of regressions is based on the four derivatives of top minus bottom temperatures and side temperatures for the left and right sides as "4-X" variables. The other set of regressions is based on the derivatives of the two side temperatures alone as "2-X" variables. The three regressions that were determined are

1. A linear regression 2. A logarithmic regression, which uses the X variables as exponents 3. The more common logarithmic form of regression, which uses the X variables as bases.

Equations used in obtaining power law regression estimates are set out below in Table 4 and correlating charts are reproduced in FIGS. 19-33.

Table 4. Power Law Regressions Using Transient Data. Estimates are obtained in under 30 minutes.

Static Correlations

Schedule 20 - Static

P=7.697.dT.sub.z-A.sup.0.6167 dT.sub.h-A.sup.3.028.dT.sub.z-B.sup.0.0218.dT.sub.h-B.sup.0.834 If Pressure .gtoreq.60 psia, then: P=42.06.dT.sub.z-A -110.0.dT.sub.z-B +35.15.dT.sub.z-B +9.408.dT.sub.h-B +267.9

Schedule 40 - Static p1 P=8.533.dT.sub.z-A.sup.0.2660.dT.sub.h-A.sup.-3.905.dT.sub.z-B.sup.0.1205.dT.sub.h-B.sup.-1.857

If Pressure .gtoreq.60 psia, then:

P=37.70.dT.sub.z-A -155.6.dT.sub.h-A -13.64.dT.sub.z-B -11.64.dT.sub.h-B +410.2

Table 4. (Continued) Power Law Regressions Using Transient Data. Estimates are obtained in under 30 minutes.

Schedule 80 - Static

P=7.711.dT.sub.z-A.sup.-0.0962.dT.sub.h-A.sup.-4.103. dT.sub.z-B.sup.-0.0345. dT.sub.h-B.sup.-1.289 If Pressure .gtoreq.60 psia, then: P=2958.dT.sub.z-A -62.26.dT.sub.h-A -30.56.dT.sub.z-B -133.6dT.sub.h-B -470.3 Where: z=Top-to-bottom temperature derivative A=Cooler of the two side arrays B=Hotter of the two side arrays h=Side temperature derivative. Flow Correlators

schedule 40- Flow

Pressure:

P=4.965.dT.sub.z-A.sup.-0.7931.dT.sub.h-A.sup.-1.859. dT.sub.z-B.sup.-0.3804.dT.sub.h-B.sup.-0.0185 If Pressure.gtoreq.60 psia, then: P=146.4.dT.sub.z-A -107.6.dT.sub.h-A -72.89.dT.sub.z-B +82.82.dT.sub.h-B +169.7 Velocity: V=0.0967.dT.sub.z-A.sup.-0.9370.dT.sub.h-A.sup.-0.4008.dT.sub.z-B.sup.-0.4073.dT.sub.h-B.sup.-1.030 If Velocity .gtoreq.6 fps, then: V=-14.68.dT.sub.z-A.sup.-7525.dT.sub.h-A +9.058.dT.sub.z-B -0.7625.dT.sub.h-B +1854 Schedule 80 - Flow

Pressure:

P=5.311.dT.sub.z-A.sup.-0.2313.dT.sub.h-A.sup.-0.9241.dT.sub.z-B.sup.-0.2255.dT.sub.h-B.sup.-3.855 If Pressure.gtoreq.60 psia, then: P=228.0.dT.sub.z-A +127.7.dT.sub.h-A -1575.dT.sub.z-B 242.6.dT.sub.h-B +283.8 Table 4. (Continued) Power Law Regressions Using Transient Data. Estimates are obtained in under 30 minutes.

Velocity:

V=0.3436.dT.sub.z-A.sup.-0.1120.dT.sub.h-A.sup.-1.619.dT.sub.z-B.sup.-0.3073.dT.sub.h-B.sup.1.499 If Velocity.gtoreq.6 fps, then: V=-2.470.dT.sub.z-A -17.87.dT.sub.h-A +9.416.dT.sub.z-B +9.199.dT.sub.h-B +11.41 Where: z=Top-to-bottom temperature derivative A=Cooler of the two side arrays B=Hotter of the two side arrays h=Side temperature derivative.

The results of the various regressions are shown in FIGS. 19 through 30. The 2-X variable regressions appear to be only slightly less accurate models than the 4-X variable model. Additionally, the linear regression works well at the high pressures, and the logarithmic regression works better at the lower pressures. Therefore, an initial estimate with the logarithmic regression is made first, and if the estimate is above 60 psia, then the pressure estimate is recalculated using the linear regression. For better reliability, the 2-X variable regressions are used in both cases.

Curves have been developed for the regression coefficients as a function of pipe wall thickness. FIG. 31 is an example of the 4-X logarithmic regression. Because of the apparent nonlinear relations, no regression of the coefficients was performed. If coefficients are needed for an intermediate thickness, a linear interpolation can be used.

A total of 48 flow tests were run at Cooper Energy Services from May to July, 1995. The number of tests were divided equally between Schedule-40 (Table 2) and Schedule-80 (Table 3) test sections. All data were included in the analysis except those data from runs that had known problems. The average pressure and velocities were calculated for each test run. References to Schedule 20, Schedule 40, etc. pipe in this context refers to conventional terminology in the industry where a particular schedule designation identifies a standard diameter, wall thickness, etc.

The reduced data were used to develop initial correlation curves for both the Schedule-40 and Schedule-80 data. The estimated velocity and pressure versus measured velocity and pressure for the Schedule-40 data are shown in FIGS. 32 and 33, respectively. Note that the static data are not included in these estimates.

In summary, the procedural steps for making the appropriate determination useful in the field are as follows:

Liquid-Filled Versus Gas-Filled

First measure the nominal pipe wall temperature, and then turn on the heater 12. After 10 minutes, check that the downstream (or upstream) temperatures at 22 are significantly different from zero. If the temperature rise is not at least 5.degree. F. and the top-to-bottom temperature difference at 20 and 24 is approximately zero, then the pipe is liquid-filled. If the temperature rise is greater then 5.degree. F. or the top-to-bottom temperature difference is greater than zero, then the line is gas-filled.

Flowing Versus Static Gas

After 20 minutes has passed, compare the upstream temperature at 16 with the downstream temperature at 22. If they are equal (within .+-.5.degree.), the gas is static. If the downstream temperature is at least 5.degree. F. greater than the upstream temperature, then the gas is flowing. The direction of flow is from the lower temperature to the higher temperature. Use the appropriate procedure below to estimate gas pressure.

Pressure Estimate for Static Gas

After 30 minutes measure the sidewall temperature at 22. Use the correlation in FIG. 11 for 30 minutes to estimate pressure from the measured temperature. If the pressure is near a boundary that separates the ranges that define operational procedures, the operator may wish to allow the test to proceed. After 40 minutes, again measure the sidewall temperature and estimate pressure using the correlation for 40 minutes.

Pressure Estimate For Flowing Gas

Within 30 minutes the rate of temperature rise should have peaked and have begun to tail off. Measure the peak heights in the rates of rise of the downstream sidewall temperature, T.sub.h, of FIG. 16, and the downstream top-to-bottom temperature difference, .DELTA.T.sub.z, FIG. 15. The regressions stated above are then used to calculate estimates of both pressure and velocity.

Having thus described the invention in its preferred embodiment, it will be clear that modifications may be made to the apparatus and the procedure for mounting the same without departing from the spirit of the invention. It is not the intention of the inventors to limit the invention by the drawings, nor the words used to describe the same. Rather it is intended that the invention be limited only by the scope of the appended claims.

Claims

1. A process for determining a plurality of physical characteristics of a fluid of unknown composition inside a pipe without contacting said fluid directly, said pipe having an axis, a top, a bottom, and two sides, the pipe being of unknown volume, said process comprising:

(a) measuring the temperature of the exterior surface of the pipe at a first time at a first location;

(b) continuously applying heat to the exterior surface of the pipe at a second location spaced from the first location;

(c) measuring the temperature at a second time at the first location;

(d) using the difference between the temperatures measured at the first location to determine whether the fluid is a liquid or a gas;

(e) measuring the temperature of the exterior surface of the pipe at a third location at the first time and the second time, the first and third locations being equally and axially spaced from the second location;

(f) using the difference between the second temperature measured at the first location and the second temperature measured at the third location to determine at least whether the fluid is static;

(g) determining the difference between the second time and the first time to determine a length of time between the first and second measuring steps; and

(h) comparing the difference between the second temperature measured at the first location and the first temperature measured at the first location and the length of time with data from another source to determine the pressure of the fluid.

2. A process for determining a plurality of physical characteristics of a fluid of unknown composition inside a pipe without contacting said fluid directly, said pipe having an axis, a top, a bottom, and two sides, the pipe being of unknown volume, said process comprising:

(a) measuring the temperature of the exterior surface of the pipe at a first time at a first location;

(b) continuously applying heat to the exterior surface of the pipe at a second location spaced from the first location;

(c) measuring the temperature at a second time at the first location;

(d) using the difference between the temperatures measured at the first location to determine whether the fluid is a liquid or a gas;

(e) measuring the temperature of the exterior surface of the pipe at a third location at the first time and the second time, the first and third locations being equally and axially spaced from the second location;

(f) using the difference between the second temperature measured at the first location and the second temperature measured at the third location to determine at least whether the fluid is flowing;

(g) in a plane extending generally perpendicular to the axis and passing through the first location, continuously measuring the temperatures of the exterior surface of the pipe at the top and the bottom;

(h) continuously calculating the temperature difference between the top and the bottom and the rate of rise of the temperature difference between the top and bottom to determine a peak rate of rise of the temperature difference between the top and the bottom; and

(i) comparing the peak rate of rise with data from another source to determine the velocity of the fluid.

3. A process for determining a plurality of physical characteristics of a fluid according to claim 2, further comprising:

(a) continuously measuring the temperatures of the exterior surface of the pipe at the first location;

(b) continuously calculating the rate of rise of the temperature at the first location to determine a peak rate of rise of the temperature at the first location; and

(c) comparing the peak rates of rise with data from another source to determine the pressure of the fluid.

4. A process for determining a plurality of physical characteristics of a fluid of unknown composition inside a pipe without contacting said fluid directly, said pipe having an axis, a top, a bottom, and two sides, the pipe being of unknown volume, said process comprising:

(a) measuring the temperature of the exterior surface of the pipe at a first time at a first location;

(b) continuously applying heat to the exterior surface of the pipe at a second location spaced from the first location;

(c) measuring the temperature at a second time at the first location;

(d) using the difference between the temperatures measured at the first location to determine whether the fluid is a liquid or a gas;

(e) measuring the temperature of the exterior surface of the pipe at a third location at the first time and the second time, the first and third locations being equally and axially spaced from the second location;

(f) using the difference between the second temperature measured at the first location and the second temperature measured at the third location to determine at least whether the fluid is flowing;

(g) in a plane extending generally perpendicular to the axis and passing through the first location, continuously measuring the temperatures of the exterior surface of the pipe at the top and the bottom;

(h) continuously calculating the temperature difference between the top and the bottom and the rate of rise of the temperature difference between the top and bottom to determine a peak rate of rise of the temperature difference between the top and the bottom;

(i) continuously measuring the temperatures of the exterior surface of the pipe at the first location;

(j) continuously calculating the rate of rise of the temperature at the first location to determine a peak rate of rise of the temperature at the first location; and

(k) comparing the peak rates of rise with data from another source to determine at least one of the following: the pressure of the fluid and the velocity of the fluid.