March
6, 2001
Dear Ms. Wan:
I know that you want Caltrans to approve the best
scenic railing that meets current safety standards. The Engineering
Service Center of Caltrans has steadfastly refused to approve the Wyoming
Rail as originally designed and successfully crash tested. In so doing,
it is denying the Commission and the California public the best scenic
railing that meets current safety standards.
As explained below, the Engineering Center's refusal
to accept the Wyoming does not conform to prevailing professional
methodology for determining the safety of bridge railings. Additionally,
I learned recently that in 1992 Caltrans approved and began using a
Wyoming Rail design in California. Yet, despite all of the focus on
the Wyoming Rail in Commission proceedings on the Noyo Bridge and in the
deliberations about scenic rails, Caltrans has never given the
Commission the slightest indication that it has previously approved and
used a version of the Wyoming Rail. This omission raises serious
questions that need to be answered.
The Commission does not have the capacity to
critically review the submissions that Caltrans has been making to
Commission, but the public interest requires it. I have, therefore,
written to the Director of Caltrans requesting that he perform an
independent review of the position taken by the Engineering Service Center
on the acceptability of the Wyoming Rail for use in California. I ask
that the Commission also write Mr. Morales requesting such a review.
The Engineering Center's current rejection of the
Wyoming Rail will not withstand an impartial independent review for the
following reasons:
·
The latest "safety standard" asserted by the Engineering
Center for rejecting the Wyoming Rail is not considered relevant to
judging the safety performance of bridge railings by any nationally
recognized safety authority.
·
Crash-test results as specified in NCHRP Report 350 are
considered by all nationally recognized safety authorities as the sole
determinant of the safety performance of bridge railings. The latest
version of the Wyoming Rail has achieved a safety performance (Test Level
4, or TL-4) equal to or better than all of the other scenic rails being
proposed to the Commission by Caltrans.
Further, Caltrans needs to explain why it has never
told the Coastal Commission that in 1992 it approved and began using a
Wyoming Rail design in California.
Consider each of the above in more depth.
Caltrans latest “safety standard” is not relevant to safety; crash
tests are the sole determinant of railing safety performance
In December 2000, the Engineering Service Center of
Caltrans admitted that the Wyoming Rail met the AASHTO LRFD geometric
design standards. These were the standards that it originally said the
Wyoming Rail did not meet, thus justifying its radical modification of the
crash-tested design of the Wyoming Railing. But, at the very same time it
made this admission, it created a new "safety standard" to justify its
modification of the Wyoming Rail.[i]
There is no support for Caltrans’s latest "safety
standard" in any national safety-standard organization or specification.
The American Association of State Highway Transportation Organizations
(AASHTO), the Federal Highway Administration (FHWA), and the National
Cooperative Highway Research Program (NCHRP) unanimously assert that the
only determinant of the safety rating of a railing is its
performance in crash tests.[ii]
Geometric design criteria, such as the one now being asserted by Caltrans,
are explicitly rejected as determinants of railing safety performance –
for the very good reason that crash tests have shown them to be poor
predictors of safety performance.[iii]
The assertion of a geometric design criteria to
reject the safety-tested and accepted Wyoming Rail by a professional
engineering organization (Caltrans) is particularly puzzling: AASHTO
standards in existence since the 1980’s have allowed highway departments
to use a railing that had passed crash tests even if did not meet static
loading and geometric design criteria.[iv]
As noted above, the safety community has now completely rejected geometric
design standards for judging railing safety, mandating that crash tests be
used to determine railing safety all applications.
The irrelevance of the latest Caltrans
geometric-design "safety standard" to safety performance is illustrated by
Caltrans own experience with its Type 80 railing, one of the railings it
proposes as a “scenic railing.” The first version of the Type
80 railing conformed to the latest Caltrans "safety standard," but failed
its crash test. Based on this result, Caltrans modified the design
and crash tested it again. The modified design when combined with a
sidewalk (the Type 80SW) still failed to pass the Test-Level 3 (TL-3)
crash test. As a result, the Type 80SW was only conditionally accepted by
the FHWA and by Caltrans for use where a TL-2 performance is
acceptable (traffic speeds less than 45 miles per hour). Thus, Caltrans
own approval process recognizes that crash-test results are the final
determinant of the safety performance of a railing.
The Commission should be aware that Caltrans has
proposed using the Type 80SW on the Noyo Bridge replacement. The TL-4
Wyoming Rail is indisputably safer than the TL-2 Type 80SW.
Wyoming Rail Previously Approved for Use in California
The story of the prior approval of a Wyoming Rail
design is particularly poignant because it was the outcome of a
negotiation with many parallels to the present one between the Commission
and Caltrans. The earlier negotiation took in the early 1990's between
Caltrans and the Tahoe Regional Planning Agency (TRPA). TRPA is a
congressionally mandated bi-state (Nevada/California) agency that is
organized to maintain and enhance the scenic and aesthetic qualities of
the Lake Tahoe region. The negotiation was over the railing to be used on
a 0.7 mile stretch of State Route 89 along Emerald Bay near Cascade Lake
and Eagle Falls on the southwestern shore of Lake Tahoe. Construction on
the project began in April 1992.
The publication "Flexibility in Highway Design,"
published by the FHWA in 1997, summarizes the negotiation and its results:
Although the
original barrier design proposed by Caltrans was the generally accepted
statewide standard for such applications, it was considered to be
unacceptable by the TRPA. The two-beam guardrail which was finally
installed had been crash tested by the FHWA some years earlier but had
not, at the time of this project, been adopted by Caltrans as an
acceptable guardrail type. Subsequent to the project, the two-beam
guardrail design was accepted by Caltrasn as a standard design,
particularly in areas with similar aesthetic considerations.[v]
The photo on the left below shows the railing that
Caltrans installed at Emerald Bay. The one on the right is a Wyoming Rail
(TL-3 version) near Yellowstone Park.
Caltrans Rail
on Emerald Bay, Rte 89 |
Wyoming Rail near Yellowstone |
These photos confirm that the
railing approved by Caltrans for the Emerald Bay Project is the earlier
(Test-Level 3) Wyoming Rail. Thus, in 1992 Caltrans approved the
Wyoming Rail for use in California. This rail was used in many places in
California before its use was discontinued.
The Test-Level 3 Wyoming Rail is still accepted by
the FHWA for general use on the National Highway System. It will be
fascinating to learn when and for what reasons Caltrans "decertified" its
California version of the Wyoming Rail, leaving the state with no good
aesthetic railing to use in scenic areas.
The need for a review of Caltrans
policy on scenic railings
National and state design standards give Caltrans
wide latitude to accommodate special situations. Preserving scenic values
is widely recognized as an important reason to use this flexibility. In
passing the Intermodal Surface Transportation Efficiency Act (ISTEA) of
1991, Congress emphasized, in addition to safety, the importance of
transportation design that is sensitive to its surrounding environment,
especially in historic and scenic areas.[vi]
In the case of scenic railings, Caltrans could fulfill the public’s desire
for the most visually transparent scenic railing without compromising
safety in any way. Instead the Engineering Service Center has rigidly
opposed its approval.
I urge the Commissioners to request the Director of
Caltrans to perform an independent review of the basis on which the
Engineering Service Center has denied approval of the Wyoming Rail. I
further urge you to request the Director to develop a policy guide that
make’s Caltrans’s transportation design sensitive to its surrounding
environment, especially in historic and scenic areas.
Sincerely,
Vince Taylor
Executive Director
cc: Peter Douglas (for distribution to all
Commissioners), Chris Desser, Shirley Detloff, Pedro Nava, John Woolley,
Steve Scholl
[i] This “safety standard” arbitrarily
requires that that the “opacity ratio” of a railing be higher
than that of any of the successfully crash-tested railings used to
draw Figure A1.1.1-3 of the AASHTO LRFD Standards. The opacity ratio
measures the percentage of the railing height that is covered by
railing surfaces. The higher the opacity index, the less will be the
view through the rail. Requiring a high opacity index, as Caltrans is
doing, rules out all visually transparent railings – regardless of
their tested safety performance. It is to be emphasized that the
AASHTO LRFD Standards do not contain the requirement imposed by
Caltrans.
[ii] I document this assertion in a
forthcoming publication tentatively titled “Evaluating Alternative
Scenic Railings.”
[iii] The history of the rejection of
geometric and static design criteria was summarized by J. W. Hatton of
the Federal Highway Association:
Until the late
1980’s designers relied on precedent, …AASHTO “Standard Specifications
for Highway Bridges,” and their judgment to design a bridge railing
for a particular site. The Standard Specifications, as they still do,
called for the application of a 10-Kip static load … as well as some
dimensional requirements for the openings between rail elements and
other cross section geometry [geometric design criteria]. Full-scale
crash testing was not required, although a design that “passed”
crash testing could be used even if it did not meet the static loading
and/or geometric design criteria. [Emphasis added]
In the late 1970’s
and early 1980’s, actual tests were run on several commonly-used
railings that had been designed under the static loading procedures.
The results were unexpected: several of the railings failed quite
dramatically and it was shown that static design loadings were not
sufficient to ensure adequate railing performance. As a result of
these findings, … [the FHWA] issued a policy memorandum on August 28,
1986, that stated that railings on bridges on Federal-aid projects
must be (or have been) crash tested and meet the acceptance criteria
in NCHRP Report 230 or equivalent procedures.
J. W. Hatton, Bridge Railing Design and
Testing, a Discussion with the AASHTO Highway Subcommittee on Bridges
and Structures, Technical Committee (T-7) for Guardrail and Bridge
Rail, May 14, 1996.
[v] Flexibility in Highway Design,
Federal Highway Administration, Publication Number FHWA-PD-97-062.
[i]
This “safety standard” arbitrarily requires that that the “opacity
ratio” of a railing be higher than that of any of the
successfully crash-tested railings used to draw Figure A1.1.1-3 of the
AASHTO LRFD Standards. The opacity ratio measures the percentage of
the railing height that is covered by railing surfaces. The higher
the opacity ratio, the less will be the view through the rail.
Requiring a high opacity ratio, as Caltrans is doing, rules out all
visually transparent railings – regardless of their tested safety
performance. It is to be emphasized that the AASHTO LRFD Standards
do not contain the requirement imposed by Caltrans.
[ii]
The history of the rejection of geometric and static design criteria
was summarized by J. W. Hatton of the Federal Highway Association:
Until the late 1980’s designers
relied on precedent, …AASHTO “Standard Specifications for Highway
Bridges,” and their judgment to design a bridge railing for a
particular site. The Standard Specifications, as they still do,
called for the application of a 10-Kip static load … as well as some
dimensional requirements for the openings between rail elements and
other cross section geometry [geometric design criteria]. Full-scale
crash testing was not required, although a design that “passed”
crash testing could be used even if it did not meet the static loading
and/or geometric design criteria. [Emphasis added]
In the late 1970’s and early 1980’s,
actual tests were run on several commonly-used railings that had been
designed under the static loading procedures. The results were
unexpected: several of the railings failed quite dramatically and it
was shown that static design loadings were not sufficient to ensure
adequate railing performance. As a result of these findings, … [the
FHWA] issued a policy memorandum on August 28, 1986, that stated that
railings on bridges on Federal-aid projects must be (or have been)
crash tested and meet the acceptance criteria in NCHRP Report 230 or
equivalent procedures.
J. W.
Hatton, Bridge Railing Design and Testing, a Discussion with the
AASHTO Highway Subcommittee on Bridges and Structures, Technical
Committee (T-7) for Guardrail and Bridge Rail, May 14, 1996.
[iv]
Flexibility in Highway Design, Federal Highway Administration,
Publication Number FHWA-PD-97-062.