As discussed earlier, there is no reason given in the MHI/MHRA report for the exclusion of the test results for the 60-inch anchors w/ 17-inch stabilizer plates, at the New Orleans test site. The report states that 162 anchors were tested, but results for only 160 anchors are given in the MHI/MHRA report. Taking into account the one additional test subject for Live Oak, FL, and the unreported results for two (2) anchor test subjects, this would account for the three (3) test subjects, shown in the MHI/MHRA report as Non-Applicable, for the New Orleans, LA test site. Based on previous studies of anchors, and the unreported test results, coupled with no explanation for this exclusion, it will be assumed that all attempts to install the 60-inch anchors to their full-intended depth, at the New Orleans test site, failed, or the anchors performed dismally, and were excluded. It is inappropriate to have exclude load values for 60-inch anchors with 17-inch stabilizer plates, from the New Orleans test site, without justification, or explanation. Given this, the load value for 60-inch anchors at the New Orleans test site will be considered to be zero (0) pounds, for the rest of this discussion.

In addition, we will not exclude the subsequent testing of 30-inch and 36-inch anchors w/ 12-inch stabilizer plates. This testing was accomplished with same anchor / anchor plate configuration, at two of the same test site locations, using the same test protocol. This provides for six (6) test subjects for each of these test sites, versus three (3). This is no different than adding an additional test subject for 48-inch anchors w/ 17-inch stabilizer plates at the Live Oak, FL test site, except that we are providing reasoning for the inclusion.

It is important to note before reviewing the following table, that the industry standard design load value for ALL ground anchors is based on a nominal working load of 3,150 pounds, at a 4-inch horizontal displacement, and 2-inch vertical displacement, WITHOUT FAILURE. Million of manufactured homes are anchored, based on this premise. The Federal standards require a nominal load capacity for anchoring equipment, which does not include ground anchors, of 3,150 pounds and an ultimate load capacity of 4,725 pounds. It is stated in the MHI/MHRA report that this study was not intended to validate the ultimate load capacity, or the nominal load capacity of ground anchors. However, whether the intent was there or not, the industry standard, nominal working load of 3,150 pounds for ground anchors, was once again, shown to be wishful thinking, as has been reported in previous studies. We will discuss this further, after Table 45. For the purposes of this paper only, any combined average value that exceeds 3,150 pounds, in the following table will be entered as 3,150 pounds. This does not mean that we give any credibility to the industry advertised nominal working load value, for ground anchors, of 3,150 pounds. As you will note in the table, this only occurs three (3) out of ten (10) times, which adds credibility to previous studies, which called a nominal working load of 3,150 pounds for helix soil anchors, "wishful thinking".

EXAMPLE: 30" / 12" = anchor length / stabilizer plate size

4-site average - without the two top performing sites skewing the load averages. Final averages that exceeded 3,150 pounds were entered as 3,150 pounds.

6-site average - All sites, load averages. Final averages that exceeded 3,150 pounds were entered as 3,150 pounds.

As can be seen from Table 45, the majority of the test sites fall short of the average load values proposed by the MHI/MHRA study. While the residents of New Orleans, LA and Mobile, AL would appear to have a reasonably good chance of having a home, which is adequately anchored, manufactured home occupants in Live Oak Fl, and other locations would be out of luck. This test program, if adopted, does nothing more than provide manufactured home owners/occupants an approximately 50/50 "Chance" that their home will be anchored in accordance with the federal windstorm protection standards. As has always been the case, the homebuyer will not be aware when they purchase a manufactured home, that they were also purchasing a lottery ticket for their safety, and the protection of their property.

Taking averages is common practice, and is an accepted engineering practice. However, where the public safety is at risk, and where coefficients of variations as high as 0.42 have been recorded, the coefficient of variation is factored in when calculating the recommended load values. The coefficient of variation was not factored into the values recommended by MHI/MHRA, even though variations ranging as much as 4600 pounds from the worst performing anchor, to the best performing anchor, were noted, for an anchor/stabilizer plate combination, that MHI/MHRA are recommending a load value of 2500 pounds. These high variations were apparent in all of the anchor/stabilizer plate combinations. The following figure is a visual aid, and is only intended to help you understand, the discussion of the Coefficient-of-Variation.

The Coefficient of Variation is actually a value arrived at by mathematical computation for some component, assembly, etc, and is based on the average test data values, and the range of test data results for that component, assembly, etc. Figure 1 is simply showing the obvious, and this is, in at least half of the cases, the value can be lower than the recommended value; and in the case of soil anchors, it is significantly lower than the average value recommended by MHI/MHRA. If the variation were a few hundred pounds, then the coefficient of variation would be very small, and would have little affect on the average load value, and visa-versa for large variations. A (0.30) 30% probability of failure (COV), which is the average COV from the MHI/MHRA anchor test program, for DEVICES intended to protect the user, public, and property, from an unreasonable risk of injury and/or death is not acceptable. The coefficient of variation (COV) should be used in this case, to reduce the average load capacity, to a level that would provide for a COV, that would be considered acceptable by engineering standards, for safety equipment. In the case of helix soil anchors, using the Coefficient of Variation (COV) to lower the load capacity, would probably end the use of helix soil anchors, as a single means of anchoring manufactured housing. In our opinion, this is the reason that this variable has been excluded, from factoring the load values recommended by MHI/MHRA, since it would call for anchor spacing in some instances, as close as approximately 2 3/4-feet. This would make the role that the "cones-of-influence" play, no longer a subject that could be avoided, as is currently what is being done by MHI/MHRA. (Cones-of-Influence - see original report)

Let us further discuss the supposed nominal working load value of 3,150 pounds, which MHI/MHRA did not discuss in their report. First, the MHI/MHRA anchor test method used for this study is not unique; it has been similar in all field-testing of anchors in previous studies, which attempted to validate the industry standard, 3,150-pound nominal working load for ground anchors. The result of the MHI/MHRA study supports the findings of previous studies. Ground anchors, under inclined loads, which is the primary anchor-loading configuration for manufactured home anchoring systems, perform poorly. Helix soil anchors provide low resistance to horizontal displacement, and horizontal displacement is the primary failure mode for manufactured housing in windstorms, which leads to ultimate failure.

78% of the anchors in Beaumont, TX did not meet the 3,150 pounds nominal load capacity. 75% of the anchors in Live Oak, FL did not meet the 3,150 pounds nominal load capacity. 76% of the anchors in Baxley, GA did not meet the 3,150 pounds nominal load capacity. 81% of the anchors in Edgefield, SC did not meet the 3,150 pounds nominal load capacity. 47% of the anchors in Mobile, AL did not meet the 3,150 pounds nominal load capacity. 27% (not including the 3 excluded test subjects) of the anchors in New Orleans, LA did not meet the 3,150 pounds nominal load capacity. A combined total of, 64% of the anchors did not meet the 3,150-pound nominal working load, which strengthens the argument, that the 3,150-pound nominal working load value is invalid.