9.1) Publishing Research

One of the primary ways we achieve our mission is by publishing our research findings in well respected, peer-reviewed journals. In addition to advancing the mission of the group, journal publications provide a way for students to show a body of professional work that will advance them towards graduation and establish credentials for future career opportunities. Journal publications are of paramount importance for every member of the Webber Energy Group.

In addition to journal publications, graduate students are expected to formally present their work at conferences which are attended by our peers and other professionals representing industry, academia, and government research institutions. They are excellent opportunities to network and receive critical feedback regarding your work.

Conferences will regularly require that a complete paper be submitted before you are accepted to present. Some conferences will allow you to submit just an abstract before being accepted. Conference publications may be peer-reviewed but they should not be confused with journal publications. Conference papers are not held to the same rigorous standards of journal publications and thus are not held in as great esteem on your curriculum vitae. Despite this, conference papers have a particular value in helping you establish an initial framework for future journal publications. In addition, conference publications often act as the entrance price for presenting at a conference, networking with peers, and receiving critical feedback on your research, all of which are the real value in attending a conference.

Please refer to Section 8 which provides the minimum expectations of productivity (e.g. journal publications and conference presentations) to receive a graduate degree.

9.1.1 Co-authors on Papers

 Choosing the appropriate co-authors for your paper is an important part of the process. At a minimum, you should expect to list Dr. Webber as a co-author on your technical papers due to all of the support, guidance, and intellectual contributions that he will provide on your work. However, it is possible that you will publish single author articles, such as opinion pieces in newspapers, which do not require listing Dr. Webber as a co-author. For your technical papers, it is likely you will have other co-authors in addition to Dr. Webber. In general, having co-authors is very beneficial to your work. Your peers not only provide critical feedback but often catch mistakes and lend a hand during critical times throughout graduate school. It is quite subjective for determining what justifies being added as a co-author but at a minimum the intellectual contribution should help advance the work in a material way. It is not appropriate to add group members to papers unless they contributed, but we should also happily recognize those around us that help make the paper possible. Please consult with Dr. Webber or a senior group member if you need confirmation on appropriate co-authors.

9.1.2 General Procedure for Submitting a Journal Paper 

• The first thing you should do before writing a paper is review the relevant sections of the handbook, which provides a detailed description of the writing style that is expected for the group. Appendix A also provides a checklist that should be used prior to submitting an article for review.
• Review your data in detail with your co-authors. Be sure that your conclusions are fully supported by your data.
• Outline your paper. Determine which figures you will include and be able to clearly justify the purpose of each figure. Your figures are often the most important thing in the entire paper. Many people reading your paper will often read only the abstract and the conclusion and then focus the majority of their time on just the figures. The figures should be able to tell a complete story of what you are trying to convey in the paper.
• Write up a first draft, edit it, rewrite and improve until it is acceptable to share with your co-authors. You may choose to use only the outline and figures to receive early feedback. But, please make sure that your outline is well thought out so that you are being respectful of your colleague’s time.
• Give a copy of your paper to a senior member of the group, ideally a post-doc. You should give your colleague at least two weeks to review the paper before providing you with feedback.
• Build off the feedback, make improvements and then, depending on the condition of the paper, submit a final draft to Dr. Webber. Dr. Webber should be the final review before you submit the paper to the journal.
• Depending on the journal, it might take a few weeks to a few months to receive feedback. The reviewers will generally provide comments, critiques and requested changes regarding the paper. The quality of the reviewers can vary widely and so can their comments. Do not take the comments personally if they are harsh (if they are particularly harsh or ridiculous, please let other people know so that a good laugh can be had; this is part of grad school, embrace it). Make any changes that are appropriate. Depending on the journal, you will most likely need to formally respond to all reviewer comments before the pa-per is published. Be extremely professional in all correspondence. But, you do not have to agree with every comment. Clearly state your opinion and whenever possible back it up with other third party literature, particularly if you are disagreeing with the reviewer’s comments.
•  Resubmit the revised version of your paper to the journal. The re-viewers will then review the paper again to understand the updates and changes based on their comments. They might have additional comments which will need to be addressed. This process might happen numerous times. In the worst case scenarios, it is possible that your paper might be under review like this for a year or more. Be patient.
• Your paper might be rejected. Again, don’t take it personally. In most cases the journal will let you know early if it is not of sufficient quality. If this is the case, please consult with Dr. Webber or a senior member of the group about how best to move forward. It is possible to successfully appeal a rejection. But, it can be a timely process. It might be better to take the rejection as a sign to improve the work. In other cases, the work might be of high quality but you simply got unlucky with the reviewers that were selected. It might be better for you to simply submit the paper to a different journal. Be persistent.
• If your paper is accepted you will generally go through a procedure of making sure that the text and figures are in proper format before the article goes to print. Make sure that your figures look good in both color and black and white. You will also need to complete additional paperwork, such as signing over copyright to the journal. In the end, you will hopefully have a published article that can help build a compelling body of work and advance you towards graduation and future career opportunities.

9.1.3 Choosing a Journal 

It is important to realize that all journals are not created equal – some are much better than others. The quality of a journal is often ranked by what is called an “impact factor” (IF). The IF is a measure of how many citations can be credited to papers published in that journal. This is generally used as a proxy to determine the quality of the journal – the better the journal, the higher the impact factor. The IF premise is that higher-quality work is more likely to be cited, therefore highly cited journals are higher-quality journals. Generally, you want to publish in the most highly regarded journal that you can. However, the impact factor methodology is not perfect for determining which journal is appropriate for your work. For instance, the topic of your research might be niche and involve fewer researchers worldwide. There might be a specific journal that your unique peer group submits to and reads more regularly. If that’s the case, it might be best for you to publish in the journal which is most often read by your peer group, even if that journal has a lower impact factor.

You should be aware that the broader scientific and engineering community does hold certain journals in high regard. The most prestigious journals are generally considered to be journals such as Science and Nature. Breakthroughs in fundamental science are generally required to warrant publication in the highest tier journals. However, highly impactful systems engineering or truly novel general interest articles can sometimes be accepted into these journals. There are also many extremely well regarded journals with greater focus on engineering, rather than fundamental science, such as Environmental Science and Technology and Applied Energy. Be aware that publishing in higher tier journals will often require a more intensive peer-review process. Despite the difficulty, it is often worth it considering that one of the objectives of WEG is to disseminate our findings as broadly as possible. Publishing in the highest tier of journals is something we should strive for.

Thomson Reuters maintains a database which compares journals throughout the world, including quantifiable measures such as impact factor. Thom-son Reuters publishes the results as Journal Citation Reports which is now integrated with the Web of Science.
 

Please refer to the publications listed on the WEG website for some ideas of journals to which you might consider submitting. Also, feel free to consult with Dr. Webber or a senior member of the group to help you decide the best venue for your work.

9.1.4 Revising a Journal Article

Once you have submitted a draft manuscript to a journal for review, it can take anywhere from weeks to months before you receive a response. After the journal receives your draft manuscript, the editor of the journal identifies professors, researchers, and other experts with the subject matter expertise required to review your paper. The journal sends official requests to review your paper to these individuals until a sufficient number of reviewers have been identified, typically 2–4 depending on the journal. The editor collects comments from the reviewers, and then delivers them back to you with a decision about your manuscript.

To maximize success upon resubmitting your draft manuscript to the journal, you should carefully read each of the reviewer’s comments and respond as thoroughly as possible. Do your best to identify each individual reviewer’s core concerns about the paper and then revise the manuscript to satisfy those concerns. When you resubmit the manuscript for review, you must include a letter to the editor with a point-by-point response to each reviewer’s comments that references the changes you have made to the manuscript. Table 2 presents guidelines for responding to common types of reviewer comments. In general, you should respond to the reviewers with a tone that thanks them for making your paper better, be gracious even if you are defending your work.

 Though it is possible for a paper to be accepted without any requested re-visions, it is exceedingly rare. Rather, the initial decision you receive from the journal will most likely say that the manuscript is not acceptable in its current form, but you are welcome to address each of the reviewer’s concerns and resubmit the manuscript within a specified period of time. The decision letter is followed by comments written by each of the reviewers that were assigned to your manuscript.

The example letter given below was submitted to the journal Energy with the revised version of the manuscript for the paper “A flexible model for economic operational management of grid battery energy storage.” The paper was accepted for publication on October 25, 2014.

Table 2: This table catalogues some common types of reviewer comments and guidelines for responding to each type.

Robert Fares

The University of Texas at Austin  

204 E Dean Keeton, Stop C2200 Austin TX 78712 USA

 
+1 540 272 8635

 +1 512 471 1045 fax

 robertfares@utexas.edu

 
October 16, 2014

 
Dear Dr. Lund,
 

Thank you for considering our manuscript titled “A flexible model for eco-nomic operational management of grid battery energy storage” (Ms. Ref. No.: EGY-D-14-02239). Our responses to the comments offered by the two referees are below. We carefully considered each comment and responded as thoroughly as possible. Please do not hesitate to contact us if you feel that any response below does not adequately address the referees’ comments.
 

Thank you again for your consideration and thoughtful feedback.
 

Best regards,
Robert L. Fares and Michael E. Webber

Responses to Reviewer 1:
 

1. The authors need to explain clearly how the studied battery participates the ERCOT market, since it is called a “grid battery energy storage”. Based on the parameters of the battery, it looks like the battery is pretty small. I’m wondering which market this battery can paticipate in the ERCOT. Who pays the electricity generated by the battery? Thank you for pointing out that our draft manuscript did not clearly explain how the battery we modeled could participate in the ERCOT energy market, or the size of the battery system considered.
 

To clarify how a battery system could participate in the ERCOT electricity market, we added a brief discussion of how a battery system would qualify for participation in the ERCOT market, and included a new reference ([51]). Please see lines 223–230 of the marked manuscript.
 

We also added a brief discussion explaining why we chose to model a single Li-ion battery cell. The primary reason is that the battery cell is the fundamental building block from which a larger battery system would be built. We clarify the fact that our analysis could be extended to describe any battery system using the same Li-ion battery cell as its fundamental building block. Please see lines 251–257 in the marked manuscript.
 

2. The revenue units used in the paper need to be clearly/correctly defined. kW is a unit of power, not electricity. Revenue usually should be counted by $/kWh, which is dollar per kW hour.
Thank you identifying the fact that our draft manuscript did not clearly define the revenue units used or discuss our selection of a revenue unit.
To make our analysis clearer, we added a discussion of the various normalization procedures that are typically used in the literature to illustrate the operating revenue of energy storage in a general way. Please see lines 317–325 of the marked manuscript. Furthermore, based on comments from another reviewer, we fundamentally revised our analysis to include not only revenue potential, but also battery system cost and net-present value. In doing so, we selected a new revenue unit that is more easily compared to system cost: $/MWh of energy delivered from the battery system to the grid. Please see lines 326–408 of the marked manuscript, which explain our rationale in selecting $/MWh delivered as the revenue/cost unit of interest, and then present a comprehensive NPV cost-benefit analysis.
 

3. I suggest the authors restructure the paper by moving the brief introduction of the battery, including its parameter and operation mode, the beginning of the paper.
Thank you for the suggestion. While we understand why our methods might be clearer if we restructured the paper as you suggest, we think that doing so might make one of our paper’s key contributions unclear. That is, restructuring the paper as you suggest might imply that our model is particular to the 850-mAh PL-383562 polymer Li-ion cell considered, and not generally applicable to any battery system that can be characterized by the behavioral circuit model illustrated in Figure 1.
 

We feel that the generality of the battery model selected for integration with our optimization routine is one of the paper’s key contributions. Thus, we would prefer to confine all description of the 850-mAh PL-383562 polymer Li-ion cell to Section 4.2, which demonstrates how our flexible model can be applied to a particular battery technology and system context.

Responses to Reviewer 2:
 

1. Page 1, line 43: ’5 year operating life could vary from $70–460/kW’ : kW is a unit of power not energy. The operating cost will depend on the energy stored and not the power rating. I believe that the author should consider reporting the results from kW to a more appropriate unit like kWh, kW-year or Ampere hour (at defined voltage).
 

Thank you for your suggestion. We agree that our initial choice to re-port revenue in $/kW was not clearly explained in the draft manuscript. To clarify our choice of revenue unit, we added a discussion of the various normalization procedures that are typically used in the literature to illustrate the operating revenue of energy storage in a general way. Please see lines 317–325 of the marked manuscript.
 

Moreover, based on your fifth comment, we fundamentally revised our analysis to include not only revenue potential, but also battery system cost and net-present value. In doing so, we selected a new revenue unit that is more easily compared to system cost: $/MWh of energy delivered from the battery system to the grid. Please see lines 326– 408 of the marked manuscript, which explain our rationale in selecting $/MWh delivered as the revenue/cost unit of interest, and then present a comprehensive NPV cost-benefit analysis.
 

We updated our abstract to reflect this new unit for normalized storage revenue and cost. Please see the highlighted portions of the abstract in the marked manuscript.
 

2. Page 3, line 45: The classification of battery models into electrochemical, mathematical and electrical is inaccurate. To give an example, electrochemical and electrical models are expressed in mathematical terms so they are mathematical models too. Additionally the author claims that mathematical models only use semi- empirical or analytical equations and do not use first principles equations. This is incorrect since there are lots of mathematical battery models that use fundamental first principle modelling. I suggest the author to re-evaluate the classification and reclassify battery models.
 

Thank you for pointing out the fact that our classification of battery models was incorrect. We initially selected the classification used in our draft manuscript based on those commonly used in the Electrical Engineering literature, which often distinguishes “mathematical” models from “electrical” models. After critical consideration, we agree that this distinction is inappropriate.
 

We edited our Background section to properly classify battery models as either first-principles electrochemical models or behavioral empirical models. Please see lines 62–88 and lines 156–158 in the marked manuscript.

3. Page 12, figure 2: The y axis use the ’cell power’ variable ’W’. This variable is not clearly defined before use and the authors do not explain what the discrete states -3,3 mean. The author is advised to clearly define this variable before first use and specify what each cell power state means.
 

Thank you for identifying the fact that we did not clearly define the cell power variable in the draft manuscript. The cell power variable is the real electrical power in W delivered by a single battery cell building block making up a larger integrated battery system. Positive values of cell power correspond to discharging, while negative values correspond to charging. We added a clear definition of this variable in lines 296–302 of the marked manuscript.
 

4. Page 16, line 9: As mentioned earlier in general comment #1, I do not think $/kW is the right term to use here. Cost should be dependent on energy charge/discharge and not the power rating.
Thank you once again for pointing out that we did not clearly define the selected $/kW term selected to describe storage revenue potential in the draft manuscript.
 

As discussed in our response to your first comment, we added a detailed description of the various normalization procedures that are typically used in the literature to illustrate the operating revenue and system cost of energy storage in a general way. We introduced $/kW power capacity, $/kWh energy storage capacity, and $/MWh delivered, each of which is frequently used in the literature. Please see lines 317–325 of the marked manuscript.
 

Based on the rationale of lines 326–331 in the marked manuscript, we selected $/MWh delivered as the revenue/cost term that best illustrates the NPV and materials-degradation cost of Li-ion energy storage.

5. The authors have done very detailed analysis of an extremely large scale problem (due to the 15 min time intervals used here). However in order to do a comprehensive analysis of the problem at hand it is important to include the cost of setting up the grid battery energy storage and the cost of battery degradation due to charge/discharge. This should be coupled with an NPV analysis of the investment required for setting up these grid battery storage stations. Given the amount of work the author’s have already put into this work, I think adding the NPV analysis with station setup and degration cost would be extremely beneficial and increase the impact of this work. I recommend the authors to include this extra analysis. If the authors disagree with this assessment, they should explain why they think that this analysis is not needed as a part of this work, or explain why they think this analysis would be too time consuming.
 

Thank you very much for the positive feedback on our work. We agree that an analysis of Li-ion energy storage system cost and NPV would improve the work. Thus, we added an extensive analysis showing the cost-benefit tradeoff of Li-ion energy storage used for wholesale energy arbitrage in Texas on the basis of revenue per unit of energy discharged ($/MWh delivered) and lifecycle Li-ion capital equipment, maintenance, and degradation (battery replacement) cost levelized per unit of energy discharged ($/MWh delivered). Our results show that a Li-ion battery system used to routinely provide only wholesale energy arbitrage service to ERCOT would have a very negative NPV. Please see lines 326–379 of the marked manuscript.
 

In addition to adding a NPV analysis, we reconfigured our analysis to more clearly illustrate the point we had intended to make with Table 2 in the draft manuscript: that revenue potential varies significantly from one day to the next, so it is important for the storage operator to consider when energy arbitrage is valuable and when it is not. To bet-ter illustrate this point, we created a “value-duration curve,” (Figure 7 in the marked manuscript) showing the daily revenue potential normalized to $/MWh of energy delivered observed in our analysis versus the levelized cost in $/MWh delivered reported for Li-ion battery systems in the literature. Please see lines 380–408 of the marked manuscript.