The modern lake Erie PDL mounts directly on the "Camco" orifice plate holder and requires no further connections except for one cable, used to communicate with a computer and charge the battery simultaneously. A simple four wire, quick connect, cable is normally run to surface on an ordinary wooden or PVC buoy so that data can be downloaded, the battery charged, and PDL functions changed without the need for a diver.

The modern PDL is capable of receiving additional input channels and controlling equipment such as plunger lifts, intermitters, and emergency shut down valves (ESDs). PDLs have already been used successfully in these applications.

It is essentially impossible to transmit a radio signal from lake bottom to shore and vice versa. Buoy mounted antennas would be an option, however this would still consume a sizeable amount of power and communication would be lost whenever the buoy rope snapped or the buoy became submerged (both of which happen often, particularly in winter).

The solution to the communication problem has not yet been implemented, but will rely on custom technology once again. Since the entire gathering system is composed of steel pipe (with the exception of some hoses and clamps) this pipe, together with the water, can be used to carry an electric current in the same fashion that current is applied to pipelines for cathodic protection. If alternating current is used, the frequency can be modulated (FM) and a signal can therefore be transmitted.

In a test conducted this summer, a low-frequency signal was applied to the pipeline and received offshore, confirming that communication through this means is possible.

In short, very little existing technology is available that is well suited to this lake application by meeting the requirements of low power consumption, in particular. For example, Multi brand plunger lift controllers were tested and found to consume far more power than could reasonably be supplied to the Lake floor through batteries or a small solar panel. The circuitry also cost approximately 8 times more than the circuitry employed in the Singlechips data logger.

Other dataloggers and controllers commercially available are very costly and are not set up with communications protocols that would be suitable for the lake. Radio communications would require high power and would not be suitable. No commercially available system is currently set up to communicate along the pipelines with low power and customization of commercial products would be very costly. Since no "canned" packages are available, only a major discovery could warrant the development expenditure to develop and repackage technology for this application.

Large production losses in winter can be cause for implementing the emergency response plan. It is usually necessary to fly over the lake to look for severe leaks or catastrophic failures. S large ice breaking boat may then be needed to get divers to locations and divers must go down in dangerous ice conditions to locate the source of the problem and correct it. Though networking will not enable remote repairs, it can eliminate the need to spend days to weeks searching for the source of the lost production in dangerous conditions and with a looming emergency by pinpointing the pressure or flow anomaly immediately.

Remotely actuated valves in important locations would enable shutting in of sections of the system in the case of a catastrophic failure and need to invoke emergency action. If ice conditions did not permit divers to reach the source of the venting gas for days or weeks, as has happened before, remotely actuated valves could turn off the supply of gas until the failed pipe could be serviced by divers.

Estimates by Port Colborne staff indicate that savings in quick location of lost production could easily amount to between $50,000 and $100,000. Dive costs from November to February totaled over $210,000 in 1997 and over $130,000 in 1998. Costs and lost production due to the Alma line blockage in winter 1999-2000 totaled over $500,000 (though networked PDLs would have been little help in this particular situation).

Of course it is hard to put a dollar value on missed drilling opportunities due to poor production data, but it is something to consider given the impact that success or failure of any one well can have.

It is known that there are many wells and sections of pipe that have restricted production due to water, improper sizing, solids restrictions and blockages, leaks, hydrate problems, and more. Locating these problems is virtually impossible without more data than annual flows and inspections. With better monitoring, bottlenecks and outright blockages can be located and corrective action can be taken, be it pipe replacements and diameter changes, plunger lifts, intermitters, or better pigging practices.

Remote monitoring will enable all Talisman staff to get a much better picture of flow conditions on the lake and should help to reduce doubt and discrepancy with respect to pipeline conditions and the benefits of maintenance work such as Well servicing, pigging, and leak control.

Installation of the onshore portion of the network will be nearly insignificant, likely at a total cost of less than $1000. As the network could be primarily installed during annual required well visits, diving costs to install the loggers in the Morpeth field (location selected for first network) should be minimal but could still amount to a several thousand dollars.

Installation of group metering points is the most costly part of data logger installation. This is because diving is required to rework piping to direct flow through an orifice plate. Many group-metering points have already been installed throughout the lake and the cost to install networked loggers on them should be minimal. No group metering points are yet available in the Morpeth field, but by placing dataloggers on approximately 7 big producing wells, production losses could be located quickly and gathering system pressure distribution could be effectively modeled.

Networking of dataloggers is not likely to reduce diving overall, as there will be more equipment in the lake and, realistically, some units will leak or break down otherwise and require maintenance from a diver. Redundant dive time such as for pressure & flow data should be cut however, as will production down time. In particular, diving in search of lost production, gathering production and pressure data for engineering purposes, and the majority of winter diving, boating, and flying should be reduced drastically. Many dollars are spent each year searching for lost production. Each time this happens in poor winter conditions, it must be regarded as an emergency situation.

Annual boat and diving costs run around $1.5M. Of this, over $0.5M is spent on initial well visits for the year. This number cannot be reduced unless some wells are abandoned or governmental approval can be granted for not visiting every well annually. This approval may be realistic if real-time data can be obtained from the wells to show that no problems are underway. Possibly a surface inspection would be all that is necessary to check the buoy and look for bubbles. Group meter volumes can be prorated by inline orifice tests performed on the wells every two years, as is acceptable accounting practice and built in to Field Data Capture (FDC) software.

It is thought that a large fraction of the wells in the lake could be optimized to increase production significantly. In fact, a quick review of the wells in the Morpeth system to look for plunger lift candidates resulted in just under 50% of the wells identified as having probable optimization potential through choke changes and/or plunger lift installations.

A testament to the fact that tubing water loading is a significant problem is that each year, when active well servicing used to begin, steady declines in production from the lake showed a slowing to reversing effect. Well services were not stopped due to lack of effect on production, but rather due to lack of economic success. Unloading a well can have a dramatic effect on gas rates, however that well will load up again in a time from days to hours or even minutes. Servicing 2-5 times annually at roughly $1k per service is not an economical means of keeping wellbores free of liquids. Plunger lifts on the other hand, at $10-15k per installation, can have a lasting impact on production as they will continuously de-water the wellbores, effectively servicing the well multiple times per day.

Again, it is difficult to put a dollar value on "missed opportunity due to lack of information and control", but with nearly 500 producing wells under the lake and a large fraction of those estimated to benefit from some kind of work, one can see that the potential is likely enormous.
Refer to Appendix A: Plunger Lift Case Study

Sweet gathering systems are nearly drawn down to their discharge limit, with only 30psig suction pressure at the Port Maitland Compressor. Despite this, wellbores at the back of the system are forced to produce against pressures over 100psig. Though this is not particularly high, it can be lowered through clearing of the pipeline and increase production at the station.

One way to keep the gathering system clear and also to keep the pressure low is to run an intermitter. The intermitter alternately switches different arms of the gathering system from production to shut in. This reduces the pressure in the gathering system and increases the velocity thereby more effectively sweeping the water into the station and producing the wells against lower backpressure.

A major intermitter project was installed on the Trustco line immediately after the Talisman purchase of Pembina. The project was considered a failure due to unsustainable gas flow in some locations. Poor location selection and valve timing were also problems, with boat visits to each site required to set new cycle times and collect data. The recorded flows did show, however, that the intermitter principal worked in several locations (that total production with downtime was greater than total production without the intermitter).

A project similar to the above, enhanced by networked PDLs would enable real-time switching to optimize the intermitter system and ensure no accidental or extended time shut-ins. Networked PDLs would also enable better location selection for the intermitter valves by providing better production data.

Intermitters are less likely to show great improvement in the lake than plunger lifts, but could still be economic projects in the massive gathering systems once network technology is available.

In order to correctly locate the improperly sized lines and select correct sizes to replace them with, pipeline models must be used due to the complexity of the looped gathering systems. These models require flow and pressure data, especially data that encompasses different flow rates in time that can be used to "tune" the models.

The Timesaver II offshore compressor has particularly shown the impact that pipeline restriction can have, as it could be seen from dataloggers in the Port Stanley system that one of the major legs of the system accounted for much more of the production gain than the other. The reduced pressure was transmitted along the less restricted line more effectively than the undersized line. Since this line had to be replaced due to leaks anyway, it was upsized to accommodate lower pressure gas flow more effectively.

There are also passive changes that can be made in the wellbores, facilitated by better flow and pressure data. Wellbores can of course be modeled in the same fashion that the pipelines can and chokes and tubing can be sized more effectively with less guesswork.

This well is a perfect example of a gas well that is undepleted and has formation deliverability but cannot produce due to the hydrostatic head of the water that fills the tubing. When the choke was pulled to install the plunger lift, a fluid level was encountered in the tubing about 300m above the perforations.

Logged data (see Fig. A1) from the plunger lift indicates that this well can build to over 100psi to over 600psig annulus pressure in under 10 minutes. After plunger unloading the tubing is once again loaded with fluid by the end of 30 minutes. This indicates the frivolousness of well servicing ($500-$800 each) to increase production.

It can also be seen from the logged data that the plunger is able to partially unload the well without the intermitter valve functioning, although it is less effective. This is indicated by the horizontal line on the annulus pressure in between shut in and unloading periods. It can he seen when the plunger got stuck that the annulus pressure suddenly climbed, indicating that the plunger had been partially unloading the well in between intermitter valve cycles (See Fig A2).

It can also he seen from the data that the well was able to unload through use of only the intermitter valve when the plunger was out of the well temporarily, though again not nearly as effectively. This unloading came to an end when the plant compressor was down for several hours and the well loaded up more extensively (See Fig A3).

Logged data from the plunger lift indicates that a net increment of at least 150Mcf/d should be seen at the plant, though unfortunately this amount is within routine fluctuation and fluctuations due to maintenance and construction in plant production and the sales increment could not be positively confirmed.