August 8, 2006

Andre Cayer, President
Watercycles Energy Recovery Inc.
110 Brewer St
Edenwold, SK S0G 0E0

780/628-7421, Fax: 306/771-2885

; http://www.watercycles.ca

Dr. Carmine F. Vasile

Ph.D. Electrophysics

60 Herbert Circle

Patchogue, NY 11772

(631) 758-6271 (Fax: 730-3918)

; gfxtechnology.com

Final Report Re: Defamation of Watercycles’s Products in 2007 Report

Dear Andre: Now that your application to become our ETL-Listee was approved by Intertek last week, your products can compete in all DHR markets in the US & Canada. It is therefore imperative that Zaloum’s 2007 report be quashed, on the grounds that it uses defective and inconsistent mathematical models crafted to make Power Pipe appear to outperform all competing gravity film heat exchanger (“GFX”) designs.

To help you prove the mathematical models in Zaloum’s Tables 2 & 3 are flawed and biased, I prepared Appendix I. Its Table C was created to rebut his erroneous and conflicting conclusions quoted in Footnote #[1]. The series-connected GFX’s of this Table provide about the same coiled-area as longer GFX’s of Table B.

Did you notice Fig. 12 & Table 6 reveal the cross-section your Lo-copper GFX’s are close to a Power Pipe’s, yet Zaloum failed to point out that this fact explains his conclusion: “It can be seen that in this case the GFX 40 also dominates followed by the Watercycles 36, but the order differs slightly from that of the NTU/foot”.

Tables A & C show longer versions also dominate; contrary to his false and misleading findings: “The Watercycles pipes, which did not perform very well”… They do seem to follow the same kind of pattern where the 60 inch pipes have lower values of effectiveness”. In fact, his mathematical models for Power Pipes and Retherm’s S3-60 are flawed for the same reason; they neglect lower U-factors associated with lower coil-flow velocities multi-coil GFX’s. Did you also notice Zaloum’s Table 2 has identical equations for your G3-29-LC and Retherm’s S3-60, but different equations for their effectiveness in Table 3? If this is a typo, it must be corrected.

Should you decide to file a defamation-complaint based, in part, on Zalom’s flawed mathematical models, I uploaded the attachments to my 7/29/07 email to Craig Murray (rows 3 & 5, Footnote #[2]). The error analysis of Table D clearly illustrates a strong bias in favor of Power-Pipe; perhaps because Watercycle’s Lo-cost GFX’s offer a superior Savings to Investment Ratio (SIR) than Power Pipes offering the same DHR-efficiency; something Zaloum also failed to address.

Series Connected GFX’s

Had you sent Zaloum a pair of G3-28.5-LC’s, instead of a G3-29-LC & G3-56-LC prototype, it would have made it difficult for him to justify inconsistant mathmatical models for identical GFX’s when connected in series.

Table C indicates series connected GFX’s of the same design have the same NTU-variation. Only one independent variable (x) is required for balanced flow; two (x,z) for unequal coil & drain flow rates flow rates, which is always the case in any multi-coil GFX; including the 2-6 coil Power Pipes noted in Footnote #[3].

Therefore, mathematical models based on series-connected GFX’s will increase the accuracy of SRC’s DWHR Perfomance Testing Program and simplify their proposed test procedure because a pair of short GFX’s can be used to create 4 types of GFX’s. This will lessen the burden on participants because they will have to supply only two short heat exchangers. For example, Watercycles would have to provide a pair of G3-31.5-LC’s that SRC could configure as a G3-60-LC or S3-60-LC or P3-31.5-LC, and will only have to incur the expense of deriving one accurate NTU-equation y(x,z) for a G3-31.5-LC to derive mathematical models for the other configurations.

Other Inconsistencies in Zaloum’s 2007 Report

The Power Pipe photo & invoice @ https://gfxtechnology.com/Exporting-America.pdf indicate a single-coil R3-40 Power Pipe is an infringing G3-40 made by Doucette Industries, Inc. But, according to Zaloum’s mathematical models, a GFX G3-40 passed-off as a Power Pipe R3-40 is more efficient than a Power Pipe R3-40, but an R3-23 plumbed in series with an R3-40 will be less efficient than a Power Pipe R3-60!

If you compare the photos in Fig. 8 of Zaloum’s 2006 report to those in Fig. 12 of his 2007 report, each manufacturer allows about 1.5” at each end for no-hub connectors to maximize heat transfer area. Yet Zaloum’s Table 5 indicates your 56” long prototype has 4” at each end, thereby reducing the active length (“La”) by 8”. To eliminate confusion created by Zaloum’s inconsistent model designations, Appendix I relies upon our designations: S3-29-LC for a Watercycles 30; S3-56-LC for a Watercycles 60; having coiled lengths of 26” & 53”, respectively.

Mathematical models of NTU-equations for long and short single coil GFX’s of a given design must have the same exponent when approximated by functions like those Zaloum derived by curve-fitting measured data. (That’s what I did in developing the mathematicals shown in graphs opened by the last link in Footnote #2.)

One of Zaloumn’s most glaring errors is highlighted in Table B; the use of identical NTU-equations for a Watercycles 60 and Retherm S3-60. If not a typo, this is further proof that Zaloum’s mathematical models are defective because he knows NTU & Effectiveness are related by these simple equations: ε(x,z) = y(x,z)/[1+y(x,z)] & y(x,z) = ε(x,z)/[1- ε(x,z)], with x=z for balanced flow. It’s senseless to use two sets of approximation functions for y(x,z) & ε(x,z) when one suffices for any combination of coil and drain flow rates.

· That Zaloum’s mathematical models fail to account for unequal flow rates is a fatal deficiency.

Conclusion

It should now be obvious that neither the 2006 or 2007 reports could support Gerald Van Decker’s fraudulent advertising claims in Footnote #3, or Zaloum’s defamation of Watercycle’s products. The large, unexplained errors in Table D for a Watercycles G3-56-LC -- 41% compared to only 6.7% for an R3-60 Power-Pipe -- ought to be sufficient justification for NRCan to quash Zaloumn’s 2007 report and issue a recall letter.

Yours truly,

Dr. Carmine F. Vasile,

Ph.D. Electrophysics

Appendix I

Mathematical Model Comparisons[4]

Table A

NTU-Equations [y(x)] For Short GFX’s

(Condition “B”, Balanced Flow, C_r = 1)

Model	DWV Length (L.O.A.)	Coiling Length (“La”)	NTU-Equations From Zaloum’s Table 2 (x = drain & coil flow rates in L/min)	NTU & DHR-Eff. @ 8.72 L/min (2.25 gpm)
FCCI G3-40	40”	37”	y₄₀(x) = 4.0168x^-0.6678 = (4.0168/x)(x^-0.3322)	.946 & 48.6% *50%*
Retherm C3-40	40”	37”	y_40C(x) = 3.3371x^-0.6817 = (3.3371/x)(x^-0.3183)	0.763 & 43.3%
Power Pipe R3-36	36”	33”	y₃₆(x) = 3.0514x^-0.7401 = (3.0514/x)(x^-0.2599)	0.614 & 38% *40%*
Watercycles G3-29-LC	29”	26”	Y₂₉(x) = 0.98x^-0.3465 = (0.98/x)(x^-0.6535)	0.463 & 31.6%

NOTE: If U were constant, the exponents in the above equations would all be zero; *--%* values from 2006 report; Zaloum’s model designation for Watercycle’s Lo-Copper GFX’s are confusing and illogical.

Table B

NTU-Equations For Long GFX’s (Condition “B”)

Model	L.O.A.	La	NTU-Equations From Zaloum’s Table 2	NTU & DHR-Eff. @ 8.72 L/min (2.25 gpm)
Smart Drain G3-60	60”	57”	y_Z60(x) = 4.4495x^-0.6711	1.040 & 51% *61%*
Retherm S3-60	60”	57”	y_Z60S(x) =3.1314x^-0.713	0.669 & 40% *46%*
Power Pipe R3-60	64”	61”	y_Z60(x) = 5.0866x^-0.6601	1.218 & 55% *56%*
Watercycles G3-56-LC	56”	53”	y_Z56(x) = 3.1314x^-0.713	0.669 & 40%

NOTE: *%* values from 2006 report

Table C

NTU Equations For Series Connected GFX’s

Special Short Model to be Connected in Series	L.O.A.	La	Series-Connected NTU-Equations Scaled From Table A	NTU & DHR-Eff. @ 8.72 L/min (2.25 gpm)
GFX G3-31.5	60”	57”	y₆₀(x) = 6.188x^-0.6678	1.46 & 59.3%
Retherm C3-31.5	60”	57”	y_60S(x) = N/A**	N/A
Power Pipe R3-31.5	64”	61”	y₆₀(x) = 5.640x^-0.7401	1.136 & 53.2%
Watercycles G3-31.5-LC	56”	53”	y₅₆(x) = 1.998x^-0.3465	0.944 & 48.5%

** The flow rate in each coil of an S3-60 is half the dran flow rate for Condition “B”, yet the 2007 Report fails to include mathematical models or measured data for unequal coil & flow rates required to develop accurate, multi-coil models.

Table D

Error Functions: EF[y_Bm(x)/y_Zm(x)]

Model	L.O.A.	La	Ratio of Table B Equations to corresponding NTU-Equations from Table C	1-EF(x) in % @ 8.72 L/min (2.25 gpm)
Smart Drain G3-60	60”	57”	EF₆₀(x) = 6.188x^-0.6678/4.4495x^-0.6711 = 1.391x^+.0033	-40%
Retherm S3-60	60”	57”	EF_60S(x) = N/A	N/A
Power Pipe R3-60	64”	61”	EF₆₀(x)= 5.640x^-0.7401/5.0866x^-0.6601 = 1.109x^-0.08	6.7%
Watercycles G3-56-LC	56”	53”	EF₅₆(x) = 2x^-0.3465/3.1314x^-0.713 = 0.638x^+.3665	-41%

[1]

“The Watercycles pipes, which did not perform very well, have a high vertical flow rate and their tubing wasn’t very square and more of a roundish rectangular shape with the width (longer side of the rectangle) of the tube against the pipe. The way the Watercycles pipes were manufactured, most likely by rolling the tube through a single set of rollers, caused a cavity in the centre of the tubing, creating an air pocket which impedes heat transfer flow rate, as air is a poor conductor.” (Quote from 2007 report, pg. 28)
“It can be seen that in this case the GFX 40 also dominates followed by the Watercycles 36, but the order differs slightly from that of the NTU/foot. They do seem to follow the same kind of pattern where the 60 inch pipes have lower values of effectiveness.” (Quote from 2007 report, pg. 21)

[2]

Zaloum’s 2006 report: gfxtechnology.com/NRCan-3_24_06.pdf
Zaloum’s 2007 report: https://gfxtechnology.com/NRCAN-6_29_07.pdf
My comments on the latter: https://gfxtechnology.com/Comments-2007Report.pdf
Product Switching: https://gfxtechnology.com/Exporting-America.pdf
Power Pipe Marketing Fraud: gfxtechnology.com/Fraudulent-Marketing.pdf
Denis Van Decker Fraud: https://gfxtechnology.com/VandeckerDefamation-EDU_2_05.pdf
NAHB Fraud: https://gfxtechnology.com/NAHB-Enron%20Fraud-EDU_9_00.pdf
Original GFX Specs & Drawings: https://gfxtechnology.com/contents.html#selection
Lo-Copper GFX’s made by Watercycles: https://gfxtechnology.com/Lo-Copper-GFX-Models.html
Heat Exchanger U-variation For Configuration “B” defined in 2006 report: https://gfxtechnology.com/Eff-Comp.PDF
Heat Transfer Tables & Curves for constant & flow-dependant U-values: https://gfxtechnology.com/T-C.pdf

[3] “The newly patented Power-Pipe™ drainwater heat recovery unit by RenewABILITY Energy Inc. (REI) offers improved performance over the original GFX™ designs. Added to these advantages is the simplified model line-up that makes specifying the Power-Pipe™ more straightforward in any project.

In the original GFX™ design, pressure loss was reduced by splitting the fresh water flow into two or three single coils covering different sections of the drainpipe. Efficiency was reduced, as the heat exchanger was not counter-flow, and the lengths of exposed drainpipe between the coils reduced the heat transfer area. In addition, the added complexity meant higher cost for these units, which were required in most applications.

The new Power-Pipe™ design, on the other hand, uses multiple coils wound together in parallel, giving the following advantages:

It is a true counter-flow heat exchanger and therefore more efficient
The coils take advantage of the full length of drainpipe, increasing the overall heat transfer area.
REI’s improved manufacturing process provides greater contact area between coil and drainpipe, further improving performance
Most importantly, the number of coils can be varied from 1 to 6 to meet project pressure requirements without increasing the cost of the unit. When Pressure drop is reduced to insignificant levels, then other characteristics such as increased turbulence can be introduced to further increase heat exchange efficiency.” (Quoted from gfxtechnology.com/MarketingFraud.pdf)

[4] According to the photos in Fig. 8 of the 2006 report and Fig. 12 in the 2007 report, each model allows about 1.5” for such connectors, yet Zaloum’s Table 5 appears to contradict these photos. For example, Smart Drain never made a 65” long GFX. I sent Mr. Zaloum -- at my expense -- the GFX-Star prototype (Model GS-G3-60-500) shown in the photo on pg. 2 of https://gfxtechnology.com/GFX-STAR.pdf. He refused to add it to his test program; stripped off the controller & pump then tested only the 60-inch long G3-60. I don’t know where he obtained the 65” G3-60 listed in Table 5, but I can guarantee that if I made one it would have a higher effectiveness that the 64” Power Pipe R3-60 because it would have a coiled length of 62”; not 60.5” as indicated in Table 5.